Re: [Users] migrations failing with latest master + vdsm
I have been running EL 6.4 hosts in 3.3 mode for quite some time, I only noticed this breakage in the latest master VDSM 4.13.x. Tagged vdsm versions: ovirt-3.3.0http://gerrit.ovirt.org/gitweb?p=vdsm.git;a=shortlog;h=refs/heads/ovirt-3.3.0and ovirt-3.3http://gerrit.ovirt.org/gitweb?p=vdsm.git;a=shortlog;h=refs/heads/ovirt-3.3do work (if not using master which one of these should be used with 3.3 btw?) The running version of libvirt on the hosts is: libvirt-0.10.2-18.0.1.el6_4.14.x86_64 - DHC On Tue, Oct 15, 2013 at 9:05 AM, Michal Skrivanek mskri...@redhat.comwrote: On Oct 12, 2013, at 00:30 , Dan Kenigsberg dan...@redhat.com wrote: On Fri, Oct 11, 2013 at 03:30:35PM -0500, Dead Horse wrote: VM migrations are failing with latest master engine and vdsm logs attached from engine and both hosts Hosts are EL 6.4 with latest master VDSM Thanks for your report! Thread-195::ERROR::2013-10-11 15:22:39,508::vm::304::vm.Vm::(run) vmId=`4bad94ad-c338-4ec5-8e5b-9910d58c1854`::Failed to migrate Traceback (most recent call last): File /usr/share/vdsm/vm.py, line 291, in run self._startUnderlyingMigration() File /usr/share/vdsm/vm.py, line 369, in _startUnderlyingMigration self._abortOnError else 0), AttributeError: 'module' object has no attribute 'VIR_MIGRATE_ABORT_ON_ERROR' Peter, Michal, I think that VIR_MIGRATE_ABORT_ON_ERROR is expected only in el6.5, which is still not public Bug 972675 - Fail migration when VM get paused due to EIO This must be reverted in vdsm or hacked in Engine (do not set abortOnError=True if libvirt libvirt-0.10.2-20.el6) as of http://gerrit.ovirt.org/#/c/19312/ the flag is sent for 3.3 clusters, which is correct I thought you're not supposed to have EL 6.4 host in 3.3 cluster. 6.5 should work.. Thanks, michal Additionally the hosts seem to lose connection with their storage domains (new behavior), are offline then recovered (even though not is physically wrong). ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] migrations failing with latest master + vdsm
I just took a look through the vdc_options keys and I only found key entries of AbortMigrationOnError for cluster levels 3.0, 3.1, and 3.2 but not 3.3. - DHC On Tue, Oct 15, 2013 at 10:55 AM, Dan Kenigsberg dan...@redhat.com wrote: On Tue, Oct 15, 2013 at 04:42:11PM +0200, Michal Skrivanek wrote: On Oct 15, 2013, at 16:36 , Dead Horse deadhorseconsult...@gmail.com wrote: I have been running EL 6.4 hosts in 3.3 mode for quite some time, I only noticed this breakage in the latest master VDSM 4.13.x. Tagged vdsm versions: ovirt-3.3.0 and ovirt-3.3 do work (if not using master which one of these should be used with 3.3 btw?) it is correct I suppose as a workaround you can use engine-config and set AbortMigrationOnError to false for now… the default should change to false till EL 6.5 comes out, I guess… Global configurable are evil, but yes. Please set it to False by default, until we can require a libvirt version that has it. Dan. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Cannot set disk interface in user portal
The Add disk UI in the User Portal does not populate the interface pull-down. Nor will it allow the creation of a new virtual disk as a result. The error states insufficient permissions however the user is assigned PowerUserRole at the cluster level and on the associated storage domain(s). Engine built from commit: 31cd5339cfda6ad78f5d3a1ff08ad666f734f814 2013-10-08 12:53:15,378 INFO [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (ajp--127.0.0.1-8702-10) Correlation ID: 567417ad, Job ID: 6f774854-920b-4217-9906-a484841b843a, Call Stack: null, Custom Event ID: -1, Message: VM sdfa was created by ovirttest. 2013-10-08 12:53:15,396 INFO [org.ovirt.engine.core.bll.AddVmFromScratchCommand] (ajp--127.0.0.1-8702-10) Lock freed to object EngineLock [exclusiveLocks= key: sdfa value: VM_NAME , sharedLocks= ] 2013-10-08 12:53:15,433 INFO [org.ovirt.engine.core.bll.MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type AddVmInterface invoked with no actions 2013-10-08 12:53:15,445 INFO [org.ovirt.engine.core.bll.MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type UpdateVmInterface invoked with no actions 2013-10-08 12:53:15,459 INFO [org.ovirt.engine.core.bll.MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type RemoveVmInterface invoked with no actions 2013-10-08 12:53:19,540 ERROR [org.ovirt.engine.core.bll.GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-10) Query execution failed due to insufficient permissions. 2013-10-08 12:54:29,015 WARN [org.ovirt.engine.core.bll.AddDiskCommand] (ajp--127.0.0.1-8702-10) CanDoAction of action AddDisk failed. Reasons:VAR__ACTION__ADD,VAR__TYPE__VM_DISK,VALIDATION.DISK_INTERFACE.NOT_NULL 2013-10-08 12:54:34,891 ERROR [org.ovirt.engine.core.bll.GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-10) Query execution failed due to insufficient permissions. 2013-10-08 12:56:14,937 ERROR [org.ovirt.engine.core.bll.GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-9) Query execution failed due to insufficient permissions. 2013-10-08 12:56:27,095 WARN [org.ovirt.engine.core.bll.AddDiskCommand] (ajp--127.0.0.1-8702-4) CanDoAction of action AddDisk failed. Reasons:VAR__ACTION__ADD,VAR__TYPE__VM_DISK,VALIDATION.DISK_INTERFACE.NOT_NULL - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Cannot set disk interface in user portal
Done, BZ1016826 https://bugzilla.redhat.com/show_bug.cgi?id=1016826 ( https://bugzilla.redhat.com/show_bug.cgi?id=1016826) opened. - DHC On Tue, Oct 8, 2013 at 1:36 PM, Itamar Heim ih...@redhat.com wrote: On 10/08/2013 09:17 PM, Einav Cohen wrote: it seems that the GetVirtioScsiControllers query should be marked as VdcQueryAuthType.User in the VdcQueryType enum, or user portal code has to change to not include a call to this query. @Derez - thoughts? DHC - I think warrants a bug to track in any case, please open one. thanks, Itamar - Original Message - From: Dead Horse deadhorseconsult...@gmail.com** To: users@ovirt.org users@ovirt.org Sent: Tuesday, October 8, 2013 2:06:10 PM Subject: [Users] Cannot set disk interface in user portal The Add disk UI in the User Portal does not populate the interface pull-down. Nor will it allow the creation of a new virtual disk as a result. The error states insufficient permissions however the user is assigned PowerUserRole at the cluster level and on the associated storage domain(s). Engine built from commit: 31cd5339cfda6ad78f5d3a1ff08ad6**66f734f814 2013-10-08 12:53:15,378 INFO [org.ovirt.engine.core.dal.**dbbroker.auditloghandling.** AuditLogDirector] (ajp--127.0.0.1-8702-10) Correlation ID: 567417ad, Job ID: 6f774854-920b-4217-9906-**a484841b843a, Call Stack: null, Custom Event ID: -1, Message: VM sdfa was created by ovirttest. 2013-10-08 12:53:15,396 INFO [org.ovirt.engine.core.bll.**AddVmFromScratchCommand] (ajp--127.0.0.1-8702-10) Lock freed to object EngineLock [exclusiveLocks= key: sdfa value: VM_NAME , sharedLocks= ] 2013-10-08 12:53:15,433 INFO [org.ovirt.engine.core.bll.**MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type AddVmInterface invoked with no actions 2013-10-08 12:53:15,445 INFO [org.ovirt.engine.core.bll.**MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type UpdateVmInterface invoked with no actions 2013-10-08 12:53:15,459 INFO [org.ovirt.engine.core.bll.**MultipleActionsRunner] (ajp--127.0.0.1-8702-10) MultipleActionsRunner of type RemoveVmInterface invoked with no actions 2013-10-08 12:53:19,540 ERROR [org.ovirt.engine.core.bll.**GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-10) Query execution failed due to insufficient permissions. 2013-10-08 12:54:29,015 WARN [org.ovirt.engine.core.bll.** AddDiskCommand] (ajp--127.0.0.1-8702-10) CanDoAction of action AddDisk failed. Reasons:VAR__ACTION__ADD,VAR__**TYPE__VM_DISK,VALIDATION.DISK_** INTERFACE.NOT_NULL 2013-10-08 12:54:34,891 ERROR [org.ovirt.engine.core.bll.**GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-10) Query execution failed due to insufficient permissions. 2013-10-08 12:56:14,937 ERROR [org.ovirt.engine.core.bll.**GetVirtioScsiControllersQuery] (ajp--127.0.0.1-8702-9) Query execution failed due to insufficient permissions. 2013-10-08 12:56:27,095 WARN [org.ovirt.engine.core.bll.** AddDiskCommand] (ajp--127.0.0.1-8702-4) CanDoAction of action AddDisk failed. Reasons:VAR__ACTION__ADD,VAR__**TYPE__VM_DISK,VALIDATION.DISK_** INTERFACE.NOT_NULL - DHC __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] vdsm live migration errors in latest master
Seeing failed live migrations and these errors in the vdsm logs with latest
VDSM/Engine master.
Hosts are EL6.4
Thread-1306::ERROR::2013-09-23
16:02:42,422::BindingXMLRPC::993::vds::(wrapper) unexpected error
Traceback (most recent call last):
File /usr/share/vdsm/BindingXMLRPC.py, line 979, in wrapper
res = f(*args, **kwargs)
File /usr/share/vdsm/BindingXMLRPC.py, line 211, in vmDestroy
return vm.destroy()
File /usr/share/vdsm/API.py, line 323, in destroy
res = v.destroy()
File /usr/share/vdsm/vm.py, line 4326, in destroy
response = self.releaseVm()
File /usr/share/vdsm/vm.py, line 4292, in releaseVm
self._cleanup()
File /usr/share/vdsm/vm.py, line 2750, in _cleanup
self._cleanupDrives()
File /usr/share/vdsm/vm.py, line 2482, in _cleanupDrives
drive, exc_info=True)
File /usr/lib64/python2.6/logging/__init__.py, line 1329, in error
self.logger.error(msg, *args, **kwargs)
File /usr/lib64/python2.6/logging/__init__.py, line 1082, in error
self._log(ERROR, msg, args, **kwargs)
File /usr/lib64/python2.6/logging/__init__.py, line 1082, in error
self._log(ERROR, msg, args, **kwargs)
File /usr/lib64/python2.6/logging/__init__.py, line 1173, in _log
self.handle(record)
File /usr/lib64/python2.6/logging/__init__.py, line 1183, in handle
self.callHandlers(record)
File /usr/lib64/python2.6/logging/__init__.py, line 1220, in
callHandlers
hdlr.handle(record)
File /usr/lib64/python2.6/logging/__init__.py, line 679, in handle
self.emit(record)
File /usr/lib64/python2.6/logging/handlers.py, line 780, in emit
msg = self.format(record)
File /usr/lib64/python2.6/logging/__init__.py, line 654, in format
return fmt.format(record)
File /usr/lib64/python2.6/logging/__init__.py, line 436, in format
record.message = record.getMessage()
File /usr/lib64/python2.6/logging/__init__.py, line 306, in getMessage
msg = msg % self.args
File /usr/share/vdsm/vm.py, line 107, in __str__
if not a.startswith('__')]
File /usr/share/vdsm/vm.py, line 1344, in hasVolumeLeases
if self.shared != DRIVE_SHARED_TYPE.EXCLUSIVE:
AttributeError: 'Drive' object has no attribute 'shared'
- DHC
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Designate Master Storage Domain
I should note that pausing the VM's does not seem to be an option or it is not working. Currently pausing all VM's associated with a particular storage domain then trying to place that domain into maintenance mode yields: Error while executing action: Cannot deactivate Storage. Active VMs were detected. -Please ensure all VMs associated with this Storage Domain are stopped and in the Down state first. This is with engine built from latest master. - DHC On Mon, Aug 26, 2013 at 12:08 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Live Migration is possible when the relevant storage server is not the one needing to be taken offline for maintenance. Granted a proper Gluster setup would alleviate that however when optimum performance of file backed virtual disks is a factor Gluster is not there yet. - DHC On Sun, Aug 18, 2013 at 5:47 AM, Ayal Baron aba...@redhat.com wrote: - Original Message - Pausing the VM's can work in certain situations for simple maintenance. However suppose the purpose of the storage shutdown is to move data around for certain VM's or perhaps change that particular underlying storage Then why not live migrate the relevant disks? filesystem or hardware. Thus some of the VM's will have to be down for sure however pausing them all because of the aforementioned would not be an option since it would take hours or days depending on the amount of data and the degree of change. - DHC On Fri, Aug 16, 2013 at 12:14 AM, Karli Sjöberg karli.sjob...@slu.se wrote: tor 2013-08-15 klockan 10:46 -0500 skrev Dead Horse: Itamar this is true (I have noted occasional timing issues with it actually working). But what if as the administrator I have a specific storage domain in mind that I would like to have become the master (in the case of more then two)? @Karli The idea is not to not have to shut down all the VM's or the engine just to maintenance a storage domain(s) that may happen to be on disparate storage servers . Yes I understood that. My suggestion was a workaround until such operations are possible, that we use to minimize downtime as much as possible, to pause the VM's, shut down engine, maintenance, bring engine and VM's back. Since the VM's only was paused and engine shut down, the VM's just continue going happily unknowing exactly from where they were, and a reboot of a storage takes at most 5mins, which means 5mins total of downtime in the cloud environment for that quarter, which is acceptable for just about any SLA. /Karli - DHC On Thu, Aug 15, 2013 at 9:37 AM, Itamar Heim ih...@redhat.com wrote: On 08/15/2013 06:18 AM, Dead Horse wrote: Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/ mailman/listinfo/users my understanding is you can move the storage domain A which is master to maint. engine will promote storage domain B to master and everything should continue working as is. -- Med Vänliga Hälsningar --- Karli Sjöberg Swedish University of Agricultural Sciences Box 7079 (Visiting Address Kronåsvägen 8) S-750 07 Uppsala, Sweden Phone: +46-(0)18-67 15 66 karli.sjob...@slu.se ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman
Re: [Users] so, what do you want next in oVirt?
This: http://lists.ovirt.org/pipermail/engine-devel/2013-August/005364.html and This: http://lists.ovirt.org/pipermail/users/2013-August/015713.html(freedom in mixing storage domain types no SPF Master Storage domain) - DHC On Tue, Aug 27, 2013 at 10:47 AM, Jakub Bittner j.bitt...@nbu.cz wrote: Dne 27.8.2013 17:34, Itamar Heim napsal(a): On 08/27/2013 06:27 PM, Jakub Bittner wrote: Dne 20.8.2013 23:19, Itamar Heim napsal(a): earlier in the year we did a survey for feature requests / improvements / etc. since a lot of things were added, and priorities usually change, I'd like to ask again for what do you need the most from oVirt / what are your pain points next? below[1] I've listed my understanding of what already went in from previous survey requests (to various degrees of coverage). Thanks, Itamar [1] from the top 12 V Allow disk resize V Integrate Nagios/Zabbix monitoring - via a ui plugin V Highly Available engine - via hosted engine[2] V Open vSwitch integration - via neutron integration X Allow cloning VMs without template ? Enable hypervisor upgrade/updates through engine[3] V Allow engine on an oVirt hosted VM - via hosted engine[2] V Enable guest configuration (root password, SSH keys, network) via guest agent in engine - via cloud-init X Integrate v2v into engine ? Bond/extend ovirtmgmt with a second network for HA/increased bandwidth[4] X Integrate scheduling of snapshots and VM export for backups in engine[5] V Spice – support Google Chrome - via mime based launch Other items mentioned in previous survey which should be covered by now: - Fix timeout when adding local host during all-in-one configuration - Fix engine set-up when SELinux is disabled - Provide packages for el6 (CentOS, Red Hat Enterprise Linux) - Allow multiple VMs to be deployed from the same template at the same time - ISO domains on local/GlusterS - Show IP addresses in Virtual Machines-Network Interfaces - OpenStack Quantum support (now called Neutron) - noVNC support - Support spice.html5 and websocket proxy - Add other guest OSes to list - Port oVirt guest agent to Ubuntu[6] - SLA - Allow resource time-sharing - Spice - Mac client (via mime based launch) - Spice - port XPI plug-in to Windows (not sure this will happen, but mime based launch allows using firefox now) - Spice - client for Ubuntu/Debian (should be covered via mime based launch) [2] hosted engine is in active development, but not released yet. [3] host update is supported, but not for general yum update. [4] a lot of improvements were done in this space, but i'm not sure if they cover this exact use case [5] backup api is now being pushed to master, and orchestration of backups should probably happen via 3rd part backup vendors? [6] I'm not sure packaging exists yet, but ubuntu is covered for the basic functionality of the guest agent. __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users Some time ago I was on IBM kvm virtualization preview and they show us gold image feature and I like it. It could be implemented in oVirt ;-) more info: http://pic.dhe.ibm.com/**infocenter/ibmfsb/v2r1/index.** jsp?topic=%2Fcom.ibm.**sbsolutions.doc%**2FugGoldImageOverview.htmhttp://pic.dhe.ibm.com/infocenter/ibmfsb/v2r1/index.jsp?topic=%2Fcom.ibm.sbsolutions.doc%2FugGoldImageOverview.htm Can you explain how you view this as different from ovirt templates? Next nice feature could be importing existing iSCSI domain. importing any data storage domain i assume? Yes. Possibility of direct use of HW by VM. such as? Telephone modem. PCI-express cards. Graphic cards and the absolutely fantastic feature would be to create clusters from Intel and AMD processors together! well, you can do that today if you want to via a config change. the only thing is live migration won't work (you should probably use -cpu host to get best performance, since live migration won't be used anyway) (well, in theory we could live migrate only between hosts of same cpu vendor, but not sure interesting enough use case to make cluster and scheduling more complex). though you can do that part on your own with the new pluggable scheduler, or use -cpu host to get max performance if you don't care about live migration From my point of view it is better to have slower cpu performance and possibility to use all of our servers in cluster. I would like to have live migration available from intel to amd. The problem is only in cpu instruction sets? If so, I can use only common sets. Another feature which I forgot is network between VMs and mirroring traffic. Both configurable from WUI. Thank you ;-) __**_ Users mailing list Users@ovirt.org
Re: [Users] Designate Master Storage Domain
Live Migration is possible when the relevant storage server is not the one needing to be taken offline for maintenance. Granted a proper Gluster setup would alleviate that however when optimum performance of file backed virtual disks is a factor Gluster is not there yet. - DHC On Sun, Aug 18, 2013 at 5:47 AM, Ayal Baron aba...@redhat.com wrote: - Original Message - Pausing the VM's can work in certain situations for simple maintenance. However suppose the purpose of the storage shutdown is to move data around for certain VM's or perhaps change that particular underlying storage Then why not live migrate the relevant disks? filesystem or hardware. Thus some of the VM's will have to be down for sure however pausing them all because of the aforementioned would not be an option since it would take hours or days depending on the amount of data and the degree of change. - DHC On Fri, Aug 16, 2013 at 12:14 AM, Karli Sjöberg karli.sjob...@slu.se wrote: tor 2013-08-15 klockan 10:46 -0500 skrev Dead Horse: Itamar this is true (I have noted occasional timing issues with it actually working). But what if as the administrator I have a specific storage domain in mind that I would like to have become the master (in the case of more then two)? @Karli The idea is not to not have to shut down all the VM's or the engine just to maintenance a storage domain(s) that may happen to be on disparate storage servers . Yes I understood that. My suggestion was a workaround until such operations are possible, that we use to minimize downtime as much as possible, to pause the VM's, shut down engine, maintenance, bring engine and VM's back. Since the VM's only was paused and engine shut down, the VM's just continue going happily unknowing exactly from where they were, and a reboot of a storage takes at most 5mins, which means 5mins total of downtime in the cloud environment for that quarter, which is acceptable for just about any SLA. /Karli - DHC On Thu, Aug 15, 2013 at 9:37 AM, Itamar Heim ih...@redhat.com wrote: On 08/15/2013 06:18 AM, Dead Horse wrote: Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/ mailman/listinfo/users my understanding is you can move the storage domain A which is master to maint. engine will promote storage domain B to master and everything should continue working as is. -- Med Vänliga Hälsningar --- Karli Sjöberg Swedish University of Agricultural Sciences Box 7079 (Visiting Address Kronåsvägen 8) S-750 07 Uppsala, Sweden Phone: +46-(0)18-67 15 66 karli.sjob...@slu.se ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Designate Master Storage Domain
Pausing the VM's can work in certain situations for simple maintenance. However suppose the purpose of the storage shutdown is to move data around for certain VM's or perhaps change that particular underlying storage filesystem or hardware. Thus some of the VM's will have to be down for sure however pausing them all because of the aforementioned would not be an option since it would take hours or days depending on the amount of data and the degree of change. - DHC On Fri, Aug 16, 2013 at 12:14 AM, Karli Sjöberg karli.sjob...@slu.sewrote: ** tor 2013-08-15 klockan 10:46 -0500 skrev Dead Horse: Itamar this is true (I have noted occasional timing issues with it actually working). But what if as the administrator I have a specific storage domain in mind that I would like to have become the master (in the case of more then two)? @Karli The idea is not to not have to shut down all the VM's or the engine just to maintenance a storage domain(s) that may happen to be on disparate storage servers . Yes I understood that. My suggestion was a workaround until such operations are possible, that we use to minimize downtime as much as possible, to *pause* the VM's, shut down engine, maintenance, bring engine and VM's back. Since the VM's only was paused and engine shut down, the VM's just continue going happily unknowing exactly from where they were, and a reboot of a storage takes at most 5mins, which means 5mins total of downtime in the cloud environment for that quarter, which is acceptable for just about any SLA. /Karli - DHC On Thu, Aug 15, 2013 at 9:37 AM, Itamar Heim ih...@redhat.com wrote: On 08/15/2013 06:18 AM, Dead Horse wrote: Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/ http://lists.ovirt.org/mailman/listinfo/users mailman/listinfo/users http://lists.ovirt.org/mailman/listinfo/users my understanding is you can move the storage domain A which is master to maint. engine will promote storage domain B to master and everything should continue working as is. -- Med Vänliga Hälsningar --- Karli Sjöberg Swedish University of Agricultural Sciences Box 7079 (Visiting Address Kronåsvägen 8) S-750 07 Uppsala, Sweden Phone: +46-(0)18-67 15 66 karli.sjob...@slu.se karli.sjob...@adm.slu.se ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working
I was just more curious about exactly what files/options database options/configurations in the engine had to be changed to disable SSL for this and just allow for http. I am not quite 100% on what the engine option SSLEnabled exactly disables SSL wise (EG: HTTP/VDSM?) or what effect the SSL_ONLY option in the websocket configuration has (by default it is set to false but only SSL works?). Thus I am just curious on the underpinnings and how things are tied together and cause/effect ;-) - DHC On Fri, Aug 16, 2013 at 2:42 AM, Frantisek Kobzik fkob...@redhat.comwrote: I'll try to resolve that soon. Thanks, F. - Original Message - From: Alon Bar-Lev alo...@redhat.com To: Frantisek Kobzik fkob...@redhat.com Cc: Dead Horse deadhorseconsult...@gmail.com, users users@ovirt.org Sent: Friday, August 16, 2013 9:04:09 AM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working - Original Message - From: Frantisek Kobzik fkob...@redhat.com To: Alon Bar-Lev alo...@redhat.com Cc: Dead Horse deadhorseconsult...@gmail.com, users users@ovirt.org Sent: Friday, August 16, 2013 9:58:27 AM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Hi, exactly - the fact about the vdc option is true. (and I think we also have to allow serving novnc/spice-html5 pages using plain http. afaik now apache or jboss forces you to https). No... just a setting for the proxy. As the html files them-selves comes from same location of where user is on. Can you please handle that? Regards, F. - Original Message - From: Alon Bar-Lev alo...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users users@ovirt.org, Frantisek Kobzik fkob...@redhat.com Sent: Friday, August 16, 2013 8:45:05 AM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org, Frantisek Kobzik fkob...@redhat.com Sent: Friday, August 16, 2013 3:55:28 AM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Curiously if one wanted the disable the need to download the Server CA certificate what are the changes needed to do so? (Realizing the security implications) I do not understand, what alternative do you propose? You can disable ssl but Frantisek, we need a vdc option for that so url will contain http or https. On Fri, Aug 2, 2013 at 2:49 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Friday, August 2, 2013 10:39:48 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Thanks Alon, That did the trick. Is there any way to get the engine to push this cert to a first time visitor by default? - DHC Well, it is actually depend on browser behavior... Internet Explorer does allow you to trust the root. I could not find such option in firefox. Frantisek: Maybe we can have the link for the ca certificate so people can press it to establish trust. Have you tried to perform XMLHttpRequest and see if you get some error we can use to warn user? On Fri, Aug 2, 2013 at 1:18 AM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 11:06:11 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Attached Firefox and Chrome screenshots of Certificates. errors thrown by websockify Firefox: 1: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Chrome: 11: handler exception: WSRequestHandler instance has no attribute 'last_code' For Firefox it looks like firefox needs a bit of proding to get it to accept the Websocket CA Cert: https://github.com/kanaka/websockify/issues/34 The error generated by chrome seems to be a websockify issue: https://github.com/kanaka/noVNC/issues/86 https://github.com/kanaka/websockify/issues/22#issuecomment-3263065 https://github.com/kanaka/noVNC/issues/177 In any event I got both Chrome and Firefox working by manually browsing to: https://ENGINEFQDN:6100 and accepting the self signed cert This is because your browser does not support the CA. Please go to: http://engine/ca.crt
Re: [Users] Designate Master Storage Domain
Itamar this is true (I have noted occasional timing issues with it actually working). But what if as the administrator I have a specific storage domain in mind that I would like to have become the master (in the case of more then two)? @Karli The idea is not to not have to shut down all the VM's or the engine just to maintenance a storage domain(s) that may happen to be on disparate storage servers . - DHC On Thu, Aug 15, 2013 at 9:37 AM, Itamar Heim ih...@redhat.com wrote: On 08/15/2013 06:18 AM, Dead Horse wrote: Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users my understanding is you can move the storage domain A which is master to maint. engine will promote storage domain B to master and everything should continue working as is. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Designate Master Storage Domain
+1 to Adding a Make Master Storage Domain Button ;-) On Thu, Aug 15, 2013 at 11:09 AM, Liron Aravot lara...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Itamar Heim ih...@redhat.com Cc: users@ovirt.org users@ovirt.org Sent: Thursday, August 15, 2013 6:46:00 PM Subject: Re: [Users] Designate Master Storage Domain Itamar this is true (I have noted occasional timing issues with it actually working). But what if as the administrator I have a specific storage domain in mind that I would like to have become the master (in the case of more then two)? In that case - you could reinitialize the pool using unattached domain (you chose which one) - this domain will become the new master. 1. Create new unattached domain/Detach the domain that you want to be the master from the pool (if possible) 2. Move all the pool domain to maintenance (the master should be put to maintenance last to avoid reconstructing to another domain). 3. Right click on the data center, chose reinitialize storage pool - you can chose which domain do you want to be the master from the unattached domains. The other option is manual db intervention :) Possibly we could add a possibility to manually change the master domain, same as was added for the spm (Make SPM button that was added recently). @Karli The idea is not to not have to shut down all the VM's or the engine just to maintenance a storage domain(s) that may happen to be on disparate storage servers . - DHC On Thu, Aug 15, 2013 at 9:37 AM, Itamar Heim ih...@redhat.com wrote: On 08/15/2013 06:18 AM, Dead Horse wrote: Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC __ _ Users mailing list Users@ovirt.org http://lists.ovirt.org/ mailman/listinfo/users my understanding is you can move the storage domain A which is master to maint. engine will promote storage domain B to master and everything should continue working as is. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working
Curiously if one wanted the disable the need to download the Server CA certificate what are the changes needed to do so? (Realizing the security implications) On Fri, Aug 2, 2013 at 2:49 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Friday, August 2, 2013 10:39:48 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Thanks Alon, That did the trick. Is there any way to get the engine to push this cert to a first time visitor by default? - DHC Well, it is actually depend on browser behavior... Internet Explorer does allow you to trust the root. I could not find such option in firefox. Frantisek: Maybe we can have the link for the ca certificate so people can press it to establish trust. Have you tried to perform XMLHttpRequest and see if you get some error we can use to warn user? On Fri, Aug 2, 2013 at 1:18 AM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 11:06:11 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Attached Firefox and Chrome screenshots of Certificates. errors thrown by websockify Firefox: 1: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Chrome: 11: handler exception: WSRequestHandler instance has no attribute 'last_code' For Firefox it looks like firefox needs a bit of proding to get it to accept the Websocket CA Cert: https://github.com/kanaka/websockify/issues/34 The error generated by chrome seems to be a websockify issue: https://github.com/kanaka/noVNC/issues/86 https://github.com/kanaka/websockify/issues/22#issuecomment-3263065 https://github.com/kanaka/noVNC/issues/177 In any event I got both Chrome and Firefox working by manually browsing to: https://ENGINEFQDN:6100 and accepting the self signed cert This is because your browser does not support the CA. Please go to: http://engine/ca.crt And install that certificate as trusted, remove the explicit trust you have added, and try again. Not pretty but it worked. - DHC On Thu, Aug 1, 2013 at 2:08 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 9:59:14 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working That did the trick for getting the websocket proxy configured ( i backed out all my changes prior to running engine-setup). I do notice that it still seems to leave the ovirt-websocket-proxy.conf in it's default state and makes no dedications to it. Instead it generated /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf I also noted engine setup generated: /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/keys/websocket-proxy.key.nopass /etc/pki/ovirt-engine/requests/websocket-proxy.req None the less still neither spice nor novnc will connect. I tried changing Engine:6100 to EngineIP:6100 so that IP would be used instead. However using either the FQDN or IP still yielded the same results. You should not touch anything... all should be configured... Make sure your browser trust the *CA* of the engine and not the engine certificate directly. And try to open vnc console via webadmin. There was nothing interesting in the logs either. I do notice that whilst the websocket-proxy service is running I never see an websockify processes but instead in /var/log/messages I see: Aug 1 13:44:10 ovirtfoo ovirt-websocket-proxy.py[435]: 11: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Thus I changed SSL_ONLY=True to SSL_ONLY=False in /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf and restarted engine and websocket-proxy No dice it still generated the same error as above during an attempted connection to /var/log/messages I also not the following error message at VM power off (albeit I am guessing it has nothing to do with this issue): 2013-08-01 13:41:03,742 ERROR [org.ovirt.engine.core.vdsbroker.DestroyVmVDSCommand] (pool-6-thread-50) [304efb3e] VDS::destroy
[Users] Designate Master Storage Domain
Is there any method of designating which domain should be the master storage domain or forcibly changing the role to a different storage domain? EG: Given the following example Storage Domain A (Master) -- NFS -- Storage Server 1 Storage Domain B -- NFS -- Storage Server 2 One wants to do maintenance to Storage Server 1 but in doing so the Master storage domain is hosted from Storage Server 1. Thus the net result of taking down Storage Server 1 is that one must also take down Storage Server 2. Thus we know we must shut down VM's from Storage Domain A to maintenance Storage Server 1. Suppose however that VM's are running that we don't want to shut down and are hosted from Storage Domain B via Storage Server 2. We would want to be able to promote Storage Domain B to Master so that we can take down Storage Domain A to do maintenance to Storage Server 1. Once we are done with maintenance to Storage Server 1 we can bring Storage Domain A back on line, re-designate it as Master if desired and bring it's VM's back online. I know I have seen this occur automatically to a point when a Storage Domain goes missing that is the Master Domain but I have not noted any manual method of doing so given the above scenario. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] engine branding still not working
Looks like patch: http://gerrit.ovirt.org/#/c/17484/ was merged but in latest master the CSS styles are still not being applied. Anything particular needed to make things work (i assume the default ovirt.brand should work)? - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] engine branding still not working
;-) On Fri, Aug 2, 2013 at 2:34 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Friday, August 2, 2013 10:23:42 PM Subject: [Users] engine branding still not working Looks like patch: http://gerrit.ovirt.org/#/c/17484/ was merged but in latest master the CSS styles are still not being applied. Anything particular needed to make things work (i assume the default ovirt.brand should work)? Correct. There is pending fixs[1][2], however they are vested... maybe they will be better in few days like whine. Regards, Alon [1] http://gerrit.ovirt.org/#/c/17567/ [2] http://gerrit.ovirt.org/#/c/17588/ - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working
Thanks Alon, That did the trick. Is there any way to get the engine to push this cert to a first time visitor by default? - DHC On Fri, Aug 2, 2013 at 1:18 AM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 11:06:11 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working Attached Firefox and Chrome screenshots of Certificates. errors thrown by websockify Firefox: 1: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Chrome: 11: handler exception: WSRequestHandler instance has no attribute 'last_code' For Firefox it looks like firefox needs a bit of proding to get it to accept the Websocket CA Cert: https://github.com/kanaka/websockify/issues/34 The error generated by chrome seems to be a websockify issue: https://github.com/kanaka/noVNC/issues/86 https://github.com/kanaka/websockify/issues/22#issuecomment-3263065 https://github.com/kanaka/noVNC/issues/177 In any event I got both Chrome and Firefox working by manually browsing to: https://ENGINEFQDN:6100 and accepting the self signed cert This is because your browser does not support the CA. Please go to: http://engine/ca.crt And install that certificate as trusted, remove the explicit trust you have added, and try again. Not pretty but it worked. - DHC On Thu, Aug 1, 2013 at 2:08 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 9:59:14 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working That did the trick for getting the websocket proxy configured ( i backed out all my changes prior to running engine-setup). I do notice that it still seems to leave the ovirt-websocket-proxy.conf in it's default state and makes no dedications to it. Instead it generated /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf I also noted engine setup generated: /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/keys/websocket-proxy.key.nopass /etc/pki/ovirt-engine/requests/websocket-proxy.req None the less still neither spice nor novnc will connect. I tried changing Engine:6100 to EngineIP:6100 so that IP would be used instead. However using either the FQDN or IP still yielded the same results. You should not touch anything... all should be configured... Make sure your browser trust the *CA* of the engine and not the engine certificate directly. And try to open vnc console via webadmin. There was nothing interesting in the logs either. I do notice that whilst the websocket-proxy service is running I never see an websockify processes but instead in /var/log/messages I see: Aug 1 13:44:10 ovirtfoo ovirt-websocket-proxy.py[435]: 11: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Thus I changed SSL_ONLY=True to SSL_ONLY=False in /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf and restarted engine and websocket-proxy No dice it still generated the same error as above during an attempted connection to /var/log/messages I also not the following error message at VM power off (albeit I am guessing it has nothing to do with this issue): 2013-08-01 13:41:03,742 ERROR [org.ovirt.engine.core.vdsbroker.DestroyVmVDSCommand] (pool-6-thread-50) [304efb3e] VDS::destroy Failed destroying vm fec3260c-871a-4fbe-a006-9eee4fbfbbcc in vds = 5713e5c8-6252-4bce-a3f6-bbd8e1e6eb57 : ovirtnodefoo, error = org.ovirt.engine.core.vdsbroker.vdsbroker.VDSErrorException: VDSGenericException: VDSErrorException: Failed to DestroyVDS, error = Unexpected exception - DHC On Thu, Aug 1, 2013 at 1:07 PM, Alon Bar-Lev alo...@redhat.com wrote: If you install the proxy on the engine machine you just need: # yum install ovirt-engine-websocket-proxy # engine-setup then answer yes when prompt if you like to configure websocket proxy. you can execute engine-setup again even if you already installed. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Thursday, August 1, 2013 9:01:47 PM Subject: [Users] Questions on ovirt 3.3 browser based spice/novnc working After Referencing: http://www.ovirt.org/Features/noVNC_console http://www.ovirt.org/Features/SpiceHTML5
Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working
That did the trick for getting the websocket proxy configured ( i backed out all my changes prior to running engine-setup). I do notice that it still seems to leave the ovirt-websocket-proxy.conf in it's default state and makes no dedications to it. Instead it generated /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf I also noted engine setup generated: /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/keys/websocket-proxy.key.nopass /etc/pki/ovirt-engine/requests/websocket-proxy.req None the less still neither spice nor novnc will connect. I tried changing Engine:6100 to EngineIP:6100 so that IP would be used instead. However using either the FQDN or IP still yielded the same results. There was nothing interesting in the logs either. I do notice that whilst the websocket-proxy service is running I never see an websockify processes but instead in /var/log/messages I see: Aug 1 13:44:10 ovirtfoo ovirt-websocket-proxy.py[435]: 11: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Thus I changed SSL_ONLY=True to SSL_ONLY=False in /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf and restarted engine and websocket-proxy No dice it still generated the same error as above during an attempted connection to /var/log/messages I also not the following error message at VM power off (albeit I am guessing it has nothing to do with this issue): 2013-08-01 13:41:03,742 ERROR [org.ovirt.engine.core.vdsbroker.DestroyVmVDSCommand] (pool-6-thread-50) [304efb3e] VDS::destroy Failed destroying vm fec3260c-871a-4fbe-a006-9eee4fbfbbcc in vds = 5713e5c8-6252-4bce-a3f6-bbd8e1e6eb57 : ovirtnodefoo, error = org.ovirt.engine.core.vdsbroker.vdsbroker.VDSErrorException: VDSGenericException: VDSErrorException: Failed to DestroyVDS, error = Unexpected exception - DHC On Thu, Aug 1, 2013 at 1:07 PM, Alon Bar-Lev alo...@redhat.com wrote: If you install the proxy on the engine machine you just need: # yum install ovirt-engine-websocket-proxy # engine-setup then answer yes when prompt if you like to configure websocket proxy. you can execute engine-setup again even if you already installed. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Thursday, August 1, 2013 9:01:47 PM Subject: [Users] Questions on ovirt 3.3 browser based spice/novnc working After Referencing: http://www.ovirt.org/Features/noVNC_console http://www.ovirt.org/Features/SpiceHTML5 and looking at some of the related engine code. I am still attempting to get the spice/novnc browser based consoles to work. I am working from a build from master yesterday I used to upgrade over a previous 3.3 master build from about a month back. VDSM version on host is 4.12.0 built minutes ago. I have installed and configured the websocket proxy like so: Set WebSocketProxy to engine ENGINEIP port 6100 engine-config -s WebSocketProxy=ENGINEIP:6100 /usr/share/ovirt-engine/bin/pki-enroll-pkcs12.sh --name=websocket-proxy --password=install --subject=/C=US/O=DHC/CN=ENGINEFQDN This generates: /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/requests/websocket-proxy.req However it does not generate the key that websockify wants so we do: openssl pkcs12 -in websocket-proxy.p12 -nocerts -nodes -out /etc/pki/ovirt-engine/keys/websocket-proxy.key The configuration of ovirt-websocket-proxy: PROXY_HOST=* PROXY_PORT=6100 SOURCE_IS_IPV6=False SSL_CERTIFICATE=/etc/pki/ovirt-engine/certs/websocket-proxy.cer SSL_KEY=/etc/pki/ovirt-engine/keys/websocket-proxy.key FORCE_DATA_VERIFICATION=False CERT_FOR_DATA_VERIFICATION=/etc/pki/ovirt-engine/certs/engine.cer SSL_ONLY=True TRACE_ENABLE=False TRACE_FILE= ENGINE_USR=/usr/share/ovirt-engine Install spice-html5 git clone http://anongit.freedesktop.org/git/spice/spice-html5.git mv spice-html5 /usr/share Test spice: In Webadmin UI we set create a VM, set display as spice, start it and set it's console to spice-html5. Result spice-html client opens in a new tab but does not connect. From engine.log: 2013-08-01 12:49:52,352 INFO [org.ovirt.engine.core.bll.SetVmTicketCommand] (ajp--127.0.0.1-8702-9) Running command: SetVmTicketCommand internal: false. Entities affected : ID: fec3260c-871a-4fbe-a006-9eee4fbfbbcc Type: VM 2013-08-01 12:49:52,371 INFO [org.ovirt.engine.core.vdsbroker.vdsbroker.SetVmTicketVDSCommand] (ajp--127.0.0.1-8702-9) START, SetVmTicketVDSCommand(HostName = ovirtnodefoo, HostId = 5713e5c8-6252-4bce-a3f6-bbd8e1e6eb57, vmId=fec3260c-871a-4fbe-a006-9eee4fbfbbcc, ticket=TKfzUQJLLrUI, validTime=120,m userName=admin@internal, userId=fdfc627c-d875-11e0-90f0-83df133b58cc), log id: 5d258049 2013-08-01 12:49:52,445 INFO
Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working
Attached Firefox and Chrome screenshots of Certificates. errors thrown by websockify Firefox: 1: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Chrome: 11: handler exception: WSRequestHandler instance has no attribute 'last_code' For Firefox it looks like firefox needs a bit of proding to get it to accept the Websocket CA Cert: https://github.com/kanaka/websockify/issues/34 The error generated by chrome seems to be a websockify issue: https://github.com/kanaka/noVNC/issues/86 https://github.com/kanaka/websockify/issues/22#issuecomment-3263065 https://github.com/kanaka/noVNC/issues/177 In any event I got both Chrome and Firefox working by manually browsing to: https://ENGINEFQDN:6100 and accepting the self signed cert Not pretty but it worked. - DHC On Thu, Aug 1, 2013 at 2:08 PM, Alon Bar-Lev alo...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org Sent: Thursday, August 1, 2013 9:59:14 PM Subject: Re: [Users] Questions on ovirt 3.3 browser based spice/novnc working That did the trick for getting the websocket proxy configured ( i backed out all my changes prior to running engine-setup). I do notice that it still seems to leave the ovirt-websocket-proxy.conf in it's default state and makes no dedications to it. Instead it generated /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf I also noted engine setup generated: /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/keys/websocket-proxy.key.nopass /etc/pki/ovirt-engine/requests/websocket-proxy.req None the less still neither spice nor novnc will connect. I tried changing Engine:6100 to EngineIP:6100 so that IP would be used instead. However using either the FQDN or IP still yielded the same results. You should not touch anything... all should be configured... Make sure your browser trust the *CA* of the engine and not the engine certificate directly. And try to open vnc console via webadmin. There was nothing interesting in the logs either. I do notice that whilst the websocket-proxy service is running I never see an websockify processes but instead in /var/log/messages I see: Aug 1 13:44:10 ovirtfoo ovirt-websocket-proxy.py[435]: 11: handler exception: [Errno 1] _ssl.c:1359: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca Thus I changed SSL_ONLY=True to SSL_ONLY=False in /etc/ovirt-engine/ovirt-websocket-proxy.conf.d/10-setup.conf and restarted engine and websocket-proxy No dice it still generated the same error as above during an attempted connection to /var/log/messages I also not the following error message at VM power off (albeit I am guessing it has nothing to do with this issue): 2013-08-01 13:41:03,742 ERROR [org.ovirt.engine.core.vdsbroker.DestroyVmVDSCommand] (pool-6-thread-50) [304efb3e] VDS::destroy Failed destroying vm fec3260c-871a-4fbe-a006-9eee4fbfbbcc in vds = 5713e5c8-6252-4bce-a3f6-bbd8e1e6eb57 : ovirtnodefoo, error = org.ovirt.engine.core.vdsbroker.vdsbroker.VDSErrorException: VDSGenericException: VDSErrorException: Failed to DestroyVDS, error = Unexpected exception - DHC On Thu, Aug 1, 2013 at 1:07 PM, Alon Bar-Lev alo...@redhat.com wrote: If you install the proxy on the engine machine you just need: # yum install ovirt-engine-websocket-proxy # engine-setup then answer yes when prompt if you like to configure websocket proxy. you can execute engine-setup again even if you already installed. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Thursday, August 1, 2013 9:01:47 PM Subject: [Users] Questions on ovirt 3.3 browser based spice/novnc working After Referencing: http://www.ovirt.org/Features/noVNC_console http://www.ovirt.org/Features/SpiceHTML5 and looking at some of the related engine code. I am still attempting to get the spice/novnc browser based consoles to work. I am working from a build from master yesterday I used to upgrade over a previous 3.3 master build from about a month back. VDSM version on host is 4.12.0 built minutes ago. I have installed and configured the websocket proxy like so: Set WebSocketProxy to engine ENGINEIP port 6100 engine-config -s WebSocketProxy=ENGINEIP:6100 /usr/share/ovirt-engine/bin/pki-enroll-pkcs12.sh --name=websocket-proxy --password=install --subject=/C=US/O=DHC/CN=ENGINEFQDN This generates: /etc/pki/ovirt-engine/keys/websocket-proxy.p12 /etc/pki/ovirt-engine/certs/websocket-proxy.cer /etc/pki/ovirt-engine/requests/websocket-proxy.req However it does not generate the key
[Users] vdsm v4.12.0 dependencies for el6 hosts
vdsm v4.12.0 for engine 3.3 has some unsatisfied dependencies on el6 hosts: -- Finished Dependency Resolution Error: Package: vdsm-hook-sriov-4.12.0-10.el6.noarch (el6v-test) Requires: libvirt-daemon-driver-nodedev Error: Package: vdsm-4.12.0-10.el6.x86_64 (el6v-test) Requires: mom = 0.3.2-3 Removing: mom-0.3.0-1.el6.noarch (@epel) mom = 0.3.0-1.el6 Updated By: mom-0.3.2-2.el6.noarch (el6v-test) mom = 0.3.2-2.el6 mom can be built from http://gerrit.ovirt.org/gitweb?p=mom.git but libvirt-daemon-driver-nodedev would require a newer libvirt. (https://apps.fedoraproject.org/packages/libvirt-daemon-driver-nodedev) - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] upgrade to latest master not working
Agreed, I did revert my victim VM back the pre-upgrade snapshot and retry the upgrade with the same results with yesterdays packages.I had watch on the logs but I did not see any DB upgrade failures. Given that I figured for the heck of it I would just rebuild with this mornings latest commits to see if the issue still persisted, I actually fully expected it to fail the same. However to my surprise it did not fail, thus I assume something changed between yesterday morning this morning that allowed things to work. for grins the packages I used yesterday were generated from: engine commit: ec2b57124edd5ae1fa41e986adadfefb9c5f1aa7 otopi commit: 8e3e3c27cbcd4330143d571a3ce8caa36c6d20ef ovirt-host-deply commit: 35a5b5f5e6d6c240ef1d20904bc06498b0ed81e1 and the packages I generated this morning that worked were from: engine commit: 28251d9acd1ecbe39b9b79e7e2e69342f57c9bd7 otopi commit: 10892fbf3fe805cea416dfa3ba628f98ffa88b3e ovirt-host-deply commit: e9c98c67dcdfca21091141bcb732727f50eb97d1 - DHC On Tue, Jul 30, 2013 at 3:27 PM, Doron Fediuck dfedi...@redhat.com wrote: Big thanks DHC for your time and efforts. Since neither of us managed to reproduce the issue I suggest we drop it for now. If someone else will hit into it, we should try and get more info on a possible failure which will not complete the db upgrade. Thanks again, Doron - Original Message - | From: Dead Horse deadhorseconsult...@gmail.com | To: Doron Fediuck dfedi...@redhat.com | Cc: Alon Bar-Lev alo...@redhat.com, users users@ovirt.org, Dave Chen wei.d.c...@intel.com, Eli Mesika | emes...@redhat.com | Sent: Tuesday, July 30, 2013 11:20:01 PM | Subject: Re: [Users] upgrade to latest master not working | | Ok, so something from yesterday morning to today changed. I again rebuilt | engine from the latest commit this AM. Accordingly I also built this | mornings latest commits of otopi and ovirt-host-deploy. I then attempted | upgrade and this time I did not note any major issues. The upgrade appears | to have succeeded, and engine starts and runs. | The only things that caused failures during upgrade were: | - If the previous version ovirt-engine-sdk is not removed manually and | replaced with ovirt-engine-sdk-python engine-upgrade fails due to package | dependency issues. (which then upsets ovirt-iso-uploader, | ovirt-image-uploader, ovirt-log-collector dependencies on ovirt-engine-sdk) | - If there is a leftover releasepreview directory engine-upgrade will not | run | | There were some error messages in the engine log griping about Could not | parse option AutoRecoveryAllowedTypes value. I have attached the engine | log file from the first start of the engine after upgrade. | | The other minor thing to note is that the default ovirt.brand theme is not | working. The CSS styles are not being applied. Changes to | welcome_page.template and messages.properties were picked up however. | | - DHC | | | | On Tue, Jul 30, 2013 at 11:04 AM, Doron Fediuck dfedi...@redhat.com wrote: | | | | - Original Message - | | From: Dead Horse deadhorseconsult...@gmail.com | | To: Doron Fediuck dfedi...@redhat.com | | Cc: Alon Bar-Lev alo...@redhat.com, users users@ovirt.org, | Dave Chen wei.d.c...@intel.com, Eli Mesika | | emes...@redhat.com | | Sent: Tuesday, July 30, 2013 6:55:18 PM | | Subject: Re: [Users] upgrade to latest master not working | | | | The only error I can find in the upgrade log is: | | 2013-07-29 14:06:28 DEBUG otopi.plugins.otopi.packagers.yumpackager | | yumpackager.verbose:84 Yum Done: ovirt-engine.noarch 0:3.3.0-20.fc18 - u | | 2013-07-29 14:06:28 INFO otopi.plugins.otopi.packagers.yumpackager | | yumpackager.info:88 Yum update: 10/18: ovirt-engine | | 2013-07-29 14:06:35 DEBUG otopi.plugins.otopi.packagers.yumpackager | | yumpackager.verbose:84 Yum Done: ovirt-engine.noarch 0:3.3.0-20.fc18 - u | | 2013-07-29 14:06:35 DEBUG otopi.plugins.otopi.packagers.yumpackager | | yumpackager.verbose:84 Yum Done: ovirt-engine-3.3.0-20.fc18.noarch | | 2013-07-29 14:06:37 INFO otopi.plugins.otopi.packagers.yumpackager | | yumpackager.info:88 Yum update: 11/18: ovirt-engine-dbscripts | | 2013-07-29 14:06:38 DEBUG otopi.plugins.otopi.packagers.yumpackager | | yumpackager.verbose:84 Yum Done: | ovirt-engine-dbscripts-3.3.0-20.fc18.noarch | | 2013-07-29 14:06:38 DEBUG otopi.plugins.otopi.packagers.yumpackager | | yumpackager.verbose:84 Yum Done: ovirt-engine-backend | | Traceback (most recent call last): | | File /usr/lib/python2.7/site-packages/yum/rpmtrans.py, line 464, in | | callback | | self._unInstStop( bytes, total, h ) | | File /usr/lib/python2.7/site-packages/yum/rpmtrans.py, line 564, in | | _unInstStop | | self.total_actions) | | File /usr/lib/python2.7/site-packages/otopi/miniyum.py, line 204, in | | event | | package=package.name, | | AttributeError: 'str' object has no attribute
Re: [Users] upgrade to latest master not working
The engine version I am attempting to upgrade from was built from: --- commit: dbc3d31110ce372a1fa7e06f64c4a5c64544c679 tree: ee9ee41c3e855c26f491c77fda9622c05af3fc54 parent: 82644cae97f3c546ee631ae79c925c91e7bed836 userportal,webadmin: Change remove message Change-Id: Ia934e33d1a975e0235e1a1ffae0c8a4a7af66f10 Signed-off-by: Alexander Wels aw...@redhat.com --- Thus it is version 3.3 and not 3.2. The upgrade (built from this master this AM) was attempted ovirt a fresh install of the packages built from the above commit. I also confirmed the install was fully functional. - DHC On Mon, Jul 29, 2013 at 4:08 PM, Doron Fediuck dfedi...@redhat.com wrote: - Original Message - | From: Alon Bar-Lev alo...@redhat.com | To: Doron Fediuck dfedi...@redhat.com | Cc: Dead Horse deadhorseconsult...@gmail.com, users users@ovirt.org, Dave Chen wei.d.c...@intel.com, Eli | Mesika emes...@redhat.com | Sent: Monday, July 29, 2013 11:51:35 PM | Subject: Re: [Users] upgrade to latest master not working | | | | - Original Message - | From: Doron Fediuck dfedi...@redhat.com | To: Alon Bar-Lev alo...@redhat.com | Cc: Dead Horse deadhorseconsult...@gmail.com, users | users@ovirt.org, Dave Chen wei.d.c...@intel.com, Eli | Mesika emes...@redhat.com | Sent: Monday, July 29, 2013 11:43:46 PM | Subject: Re: [Users] upgrade to latest master not working | | | | - Original Message - | | From: Alon Bar-Lev alo...@redhat.com | | To: Doron Fediuck dfedi...@redhat.com | | Cc: Dead Horse deadhorseconsult...@gmail.com, users | | users@ovirt.org, Dave Chen wei.d.c...@intel.com, Eli | | Mesika emes...@redhat.com | | Sent: Monday, July 29, 2013 11:31:42 PM | | Subject: Re: [Users] upgrade to latest master not working | | | | | | | | - Original Message - | | From: Doron Fediuck dfedi...@redhat.com | | To: Alon Bar-Lev alo...@redhat.com | | Cc: Dead Horse deadhorseconsult...@gmail.com, users | | users@ovirt.org, Dave Chen wei.d.c...@intel.com, Eli | | Mesika emes...@redhat.com | | Sent: Monday, July 29, 2013 11:27:41 PM | | Subject: Re: [Users] upgrade to latest master not working | | | | | | | | - Original Message - | | | From: Alon Bar-Lev alo...@redhat.com | | | To: Dead Horse deadhorseconsult...@gmail.com | | | Cc: users users@ovirt.org, Doron Fediuck dfedi...@redhat.com, | | | Dave | | | Chen wei.d.c...@intel.com, Eli | | | Mesika emes...@redhat.com | | | Sent: Monday, July 29, 2013 10:47:01 PM | | | Subject: Re: [Users] upgrade to latest master not working | | | | | | | | | | | | - Original Message - | | | From: Dead Horse deadhorseconsult...@gmail.com | | | To: Alon Bar-Lev alo...@redhat.com | | | Cc: users users@ovirt.org | | | Sent: Monday, July 29, 2013 10:41:34 PM | | | Subject: Re: [Users] upgrade to latest master not working | | | | | | server.log attached | | | | | | Thanks! | | | | | | My guess is that I8ce3448a[1] broke master. | | | | | | [1] http://gerrit.ovirt.org/#/c/14605/ | | | | | | | Works perfectly on my Gentoo with master. | | | | | | I doubt you are upgrading from 3.2 :) | | | | Just did: | | $ git checkout origin/engine_3.2 | $ cd backend/manager/dbscripts/ | $ ./create_db.sh -d upgrade32-test -u engine | ~/dev/ovirt/ovirt-engine/backend/manager/dbscripts | ~/dev/ovirt/ovirt-engine/backend/manager/dbscripts | Creating the database: upgrade32-test | dropdb: database removal failed: ERROR: database upgrade32-test does not | exist | user name is: engine | Creating tables... | Creating functions... | Creating common functions... | Inserting data ... | Inserting pre-defined roles ... | Running upgrade scripts... | upgrade script detected a change in Config, View or Stored Procedure... | Running upgrade sql script upgrade/pre_upgrade/_config.sql ... | | ... | | Creating stored procedures from vm_templates_sp.sql ... | Running upgrade sql script | upgrade/post_upgrade/0010_add_object_column_white_list_table.sql ... | Running upgrade sql script | upgrade/post_upgrade/0020_create_materialized_views.sql ... | Refreshing materialized views... | ~/dev/ovirt/ovirt-engine/backend/manager/dbscripts | | $ git checkout master | $ cd ../../../packaging/dbscripts/ | $ ./upgrade.sh -d upgrade32-test -u engine | upgrade script detected a change in Config, View or Stored Procedure... | Running upgrade sql script upgrade/pre_upgrade/_config.sql ... | | ... | | Creating stored procedures from vms_sp.sql ... | Running upgrade sql script | upgrade/post_upgrade/0010_add_object_column_white_list_table.sql ... | Running upgrade sql script | upgrade/post_upgrade/0020_create_materialized_views.sql ... | Refreshing materialized views... | Done. | | | Seems to work well. Am I
[Users] upgrade to latest master not working
Trying to upgrade engine built from commit: dbc3d31110ce372a1fa7e06f64c4a5c64544c679 To latest engine built from latest master running engine-upgrade results in: ***L:ERROR Internal Error: No Module named releasepreview I know I must be missing something obvious? The surrounding versions of otopi, host-deploy, etc do yum upgrade cleanly however. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] How to add internal networks
This also still works as well: http://lists.ovirt.org/pipermail/users/2012-April/007385.html - DHC On Thu, May 16, 2013 at 9:12 AM, Lior Vernia lver...@redhat.com wrote: Hello Frank, I suspect the problem is that VDSM isn't told by default to pass dummy NICs on to the engine. Have a look at the following file (exact path might vary): /usr/lib64/python2.7/site-packages/vdsm/config.py Search for a variable called fake_nics, it should exist for 3.2. By default it is empty, but you may define a pattern for fake NIC names to be passed to the engine (in your case probably dummy*). Restart the vdsm daemon and everything should be fine. Maybe you'll have to move the host to maintenance and then reactivate just to refresh the UI. Please let us know if that works. Yours, Lior. On 16/05/13 15:48, Frank Wall wrote: Hi Asaf, Am 13.05.2013 15:05, schrieb Assaf Muller: Without the ability to control the networking devices in your vicinity and setup VLANs, what you're trying to do isn't supported via oVirt. I was already guessing that it is not supported. Thanks for the official confirmation. * Create two dummy interfaces on the host, and give them static IP addresses. For example: 192.168.1.254/24, and 192.168.2.254/24. OK, I've populated the required files... /etc/modules-load.d/dummy.conf /etc/modprobe.d/dummy.conf /etc/sysconfig/network-scripts/ifcfg-dummy0 ...and after a reboot I have a new dummy NIC configured: dummy0: flags=195UP,BROADCAST,RUNNING,NOARP mtu 1500 inet 192.168.0.254 netmask 255.255.255.0 broadcast 192.168.0.255 ether 5e:61:0f:e7:f9:bc txqueuelen 0 (Ethernet) RX packets 0 bytes 0 (0.0 B) RX errors 0 dropped 0 overruns 0 frame 0 TX packets 3 bytes 210 (210.0 B) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 * Create two networks (VM) called One and Two and attach them to the dummy NICs. Unfortunately, this does not work as described. I was able to create the new VM network from the Admin Portal successfully, but I am unable to attach this network to the dummy NIC. When I choose the Setup Host Networks option from the Admin Portal: My Datacenter = My Cluster = Hosts = My Host = TAB: Network Interfaces - Setup Host Networks The dummy NIC is NOT listed, but only the physical NIC (with ovirtmgmt attached to it). Is this expected? Does oVirt suppress any non-physical NICs? Is there another way to attach the VM network to my dummy NIC? Thanks - Frank ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Feature request: allow usbtablet and vram config via UI
I just checked on VNC. In the case of a VNC console usbtablet is being used in the generated dom xml. However like VNC when a guest has no spice awareness it is in the same boat as VNC so usbtablet is helpful in getting to that point. Or just simply if it cannot be made spice aware at all as is the case for some guest types (EG: no vdagent possiblity for example). Making VRAM configurable would be a nice benefit to SPICE or VNC console. Thus allowing for a guest to have enough vram for higher resolutions and bit depths as well as more vram for 2D video operations. - DHC On Wed, Apr 3, 2013 at 2:39 PM, Itamar Heim ih...@redhat.com wrote: On 03/27/2013 05:23 PM, Dead Horse wrote: Any thoughts/feedback on this one? It really is a downer to be forced to manipulate both of these via hooks. These are both pretty standard configuration items to enhance/tweak or make various guests fully functional. will track via the bugs. the vram for spice multi monitor i see the issue with. for usb tablet, don't we always enable it for vnc display? - DHC On Fri, Mar 22, 2013 at 9:12 AM, Dead Horse deadhorseconsult...@gmail.com mailto:deadhorseconsulting@**gmail.comdeadhorseconsult...@gmail.com wrote: Allow for a usbtablet input device to be enabled and used. Perhaps via the console configuration UI for a VM in the admin and user portals. -Highly useful in the event that a guest OS does not have the spice agent loaded nor is it available for said guest OS. -Also in the case of guest OS that simply does not want to work and play well with input type='mouse' bus='ps2' - All Operating systems since ~1998 understand input type='tablet' bus='usb'/ and will deal with mouse events in absolute mode. Allow for vram and vram_ size to be configurable for CIrrus(VNC) and QXL(Spice) console types. Again probably best suited to have this on console configuration UI for a VM in the admin and user portals. - Default VRAM sizes are not enough to allow for larger resolutions and pixel depths at times. - Especially the case in multiple monitor SPICE, fullscreen, or VNC consoles - More VRAM is useful in the case of attempting to run accelerated applications within a QXL configured guest. The above can be altered via VDSM hooks however this is rather painful. Additionally these options or similar are configurable in other competing solutions. - DHC __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Template disk cloning from one storage domain to another is not working
Bug 928399 (https://bugzilla.redhat.com/show_bug.cgi?id=928399) Opened. Also opened a bug on what appears to be a related issue or a side effect of the root cause: Bug 928410 ( https://bugzilla.redhat.com/show_bug.cgi?id=928410) - DHC On Wed, Mar 27, 2013 at 1:44 AM, Daniel Erez de...@redhat.com wrote: Hi DHC, Looking at the code, it seems that the issue indeed affects specifically the UserPortal. Can you please open a new bug on ovirt-engine-userportal: https://bugzilla.redhat.com/enter_bug.cgi?product=ovirt (please attach the relevant logs to the bug) Thanks, Daniel - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Tuesday, March 26, 2013 10:46:37 PM Subject: Re: [Users] Template disk cloning from one storage domain to another is not working Attaching some log files. Also verified still not working on latest master as of today. - DHC On Mon, Mar 25, 2013 at 4:03 PM, Dead Horse deadhorseconsult...@gmail.com wrote: In further attempt to debug this I added the user as both a SuperUser and PowerUser to the entire Datacenter so the role would cascade down to all objects. I found that with the admin portal this was not an issue. However it still occurs with the UserPortal. The same issue actually occurs for the built in admin@internal user via the user portal as well. Thus this infers some interaction with the UserPortal specifically causing this. - DHC On Mon, Mar 25, 2013 at 3:33 PM, Dead Horse deadhorseconsult...@gmail.com wrote: Template disk cloning from one storage domain to another is not working correctly. Assign a user UserTemplateVM permissions to storage domain A. Create a template in storage domain A. Ensure the user has PowerUserRole permissions to another storage domain (Storage domain B). As the user create a VM from the template in storage domain A. Choose to clone the disk to Storage Domain B. The attached disk of resultant VM gets created but the disk is created in Storage domain A instead of storage domain B. This also is an artifact of the problem I originally brought up: http://lists.ovirt.org/pipermail/users/2013-March/013164.html (Still nothing) This is present in latest engine built from master with latest VDSM built from master. I noticed it also originally in the above issue. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] template provisioning permissions
Bug 928410 (https://bugzilla.redhat.com/show_bug.cgi?id=928410) opened on this issue. Additionally Bug 928399 (https://bugzilla.redhat.com/show_bug.cgi?id=928399) which is possibly related to this issue opened. - DHC On Mon, Mar 18, 2013 at 10:02 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Verified this is present in latest engine built from master with latest VDSM built from master. On the surface this literally seems as simple as a lack of Read-Only access to the template image when requesting to clone it from the template on the storage domain wherein the user cloning from the template has no permissions. - DHC On Wed, Mar 13, 2013 at 4:34 PM, Dead Horse deadhorseconsult...@gmail.com wrote: Got an interesting one here as pertaining to template permissions and provisioning. Given the following setup/situation: A cluster with a user A assigned poweruser role permissions on the cluster. - User A is assigned poweruser role permissions to storage domain A - User A is a consumer of quota A which is assigned to specific storage domain A A cluster with a user B assigned poweruser role permissions on the cluster. - User B is assigned poweruser role permissions to storage domain B - User B is a consumer of quota B which is assigned to specific storage domain B User A creates a VM and makes it a template of it with permissions of everyone as UserTemplateBasedVM. User B tries to create a VM based on the template that User A created. While the base VM profile can be created the storage provisioning encounters an issue. Via Template provisioning option with the thin provision option will fail due to the fact that User B does not have proper permissions to User A's storage domain. The symptom of this expected failure is the target storage domain pull-down is empty. (It really should show something or be greyed out rather than just be blank at least some sort of user notification). The real issue here is with the clone provisioning option. The idea here is to be to clone a copy of the template disks into User B's storage domain as a target where User B has poweruser role permissions. The problem here is that this fails just like the above thin provision which should not be the case. The target pulldown still blank it should by default show the target storage domain to which User B has permissions to that being Storage domain B. Further debugging yields that by assigning UserTemplateVM permissions to User A's storage domain allows User B to use either of the options above although the only one really desired is the clone option since we don't want User B creating VM's in User A's storage domain. There still however was an issue upon selecting clone and selecting Storage domain B as the target the VM is created but the disk is created in Storage domain A instead of storage domain B. Running build of the engine is built from commit: 7354d3283627bdbe30dd9c15ce45eba375280a8c - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Feature request: allow usbtablet and vram config via UI
Any thoughts/feedback on this one? It really is a downer to be forced to manipulate both of these via hooks. These are both pretty standard configuration items to enhance/tweak or make various guests fully functional. - DHC On Fri, Mar 22, 2013 at 9:12 AM, Dead Horse deadhorseconsult...@gmail.comwrote: Allow for a usbtablet input device to be enabled and used. Perhaps via the console configuration UI for a VM in the admin and user portals. -Highly useful in the event that a guest OS does not have the spice agent loaded nor is it available for said guest OS. -Also in the case of guest OS that simply does not want to work and play well with input type='mouse' bus='ps2' - All Operating systems since ~1998 understand input type='tablet' bus='usb'/ and will deal with mouse events in absolute mode. Allow for vram and vram_ size to be configurable for CIrrus(VNC) and QXL(Spice) console types. Again probably best suited to have this on console configuration UI for a VM in the admin and user portals. - Default VRAM sizes are not enough to allow for larger resolutions and pixel depths at times. - Especially the case in multiple monitor SPICE, fullscreen, or VNC consoles - More VRAM is useful in the case of attempting to run accelerated applications within a QXL configured guest. The above can be altered via VDSM hooks however this is rather painful. Additionally these options or similar are configurable in other competing solutions. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Template disk cloning from one storage domain to another is not working
Template disk cloning from one storage domain to another is not working correctly. Assign a user UserTemplateVM permissions to storage domain A. Create a template in storage domain A. Ensure the user has PowerUserRole permissions to another storage domain (Storage domain B). As the user create a VM from the template in storage domain A. Choose to clone the disk to Storage Domain B. The attached disk of resultant VM gets created but the disk is created in Storage domain A instead of storage domain B. This also is an artifact of the problem I originally brought up: http://lists.ovirt.org/pipermail/users/2013-March/013164.html (Still nothing) This is present in latest engine built from master with latest VDSM built from master. I noticed it also originally in the above issue. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Feature request: allow usbtablet and vram config via UI
Allow for a usbtablet input device to be enabled and used. Perhaps via the console configuration UI for a VM in the admin and user portals. -Highly useful in the event that a guest OS does not have the spice agent loaded nor is it available for said guest OS. -Also in the case of guest OS that simply does not want to work and play well with input type='mouse' bus='ps2' - All Operating systems since ~1998 understand input type='tablet' bus='usb'/ and will deal with mouse events in absolute mode. Allow for vram and vram_ size to be configurable for CIrrus(VNC) and QXL(Spice) console types. Again probably best suited to have this on console configuration UI for a VM in the admin and user portals. - Default VRAM sizes are not enough to allow for larger resolutions and pixel depths at times. - Especially the case in multiple monitor SPICE, fullscreen, or VNC consoles - More VRAM is useful in the case of attempting to run accelerated applications within a QXL configured guest. The above can be altered via VDSM hooks however this is rather painful. Additionally these options or similar are configurable in other competing solutions. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] 3.2 final and status of spice console in ie
I have seen this one before.Try adding the your ovirt manager to the Internet Explorer trusted sites. -DHC On Mon, Mar 18, 2013 at 7:01 AM, Itamar Heim ih...@redhat.com wrote: On 03/18/2013 01:56 PM, Juan Jose wrote: Hello Itamar and everybody, I have tested this Howto with a Windows 7 Enterprise edition client and IE 9 and when you click over console button in User portal virt-viewer is launched but after one second it close windows and don't connect to virtual machine. I have oVirt 3.1 and I'm thinking in migrate my system to oVirt 3.2 but it will be nice to know which it is the reason because it doesn't work. Some indications about what would be the problem will be wellcome. If it is necessary some log I can provide it. hard to help without logs (spice client to begin with). will be easier helping/patching you on 3.2 of course. Many thanks, Juanjo. On Fri, Mar 15, 2013 at 10:37 AM, Itamar Heim ih...@redhat.com mailto:ih...@redhat.com wrote: On 02/28/2013 01:34 PM, Karli Sjöberg wrote: But to me one question arise, with a Windows clients, XP or 7, any version, it's possible to use the User Portal to be able users lauch the VM console?, or it's mandatory to have Fedora clients with spice-xpi for firefox. It would seem impossible with oVirt-3.1, which is the biggest reason why we cannot take into any real production at the moment:( What good does it do that it works with Fedora when the total of people using it counts to 1(me)? 1. there are wiki's on how to add spice windows support to ovirt: http://wiki.ovirt.org/wiki/**How_to_Connect_to_SPICE_** Console_With_Portalhttp://wiki.ovirt.org/wiki/How_to_Connect_to_SPICE_Console_With_Portal 2. there are patches for various solutions to this coming (mime based launched, spice.html5, novnc, maybe spice-xpi for firefox windows) HTH, Itamar __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] template provisioning permissions
Verified this is present in latest engine built from master with latest VDSM built from master. On the surface this literally seems as simple as a lack of Read-Only access to the template image when requesting to clone it from the template on the storage domain wherein the user cloning from the template has no permissions. - DHC On Wed, Mar 13, 2013 at 4:34 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Got an interesting one here as pertaining to template permissions and provisioning. Given the following setup/situation: A cluster with a user A assigned poweruser role permissions on the cluster. - User A is assigned poweruser role permissions to storage domain A - User A is a consumer of quota A which is assigned to specific storage domain A A cluster with a user B assigned poweruser role permissions on the cluster. - User B is assigned poweruser role permissions to storage domain B - User B is a consumer of quota B which is assigned to specific storage domain B User A creates a VM and makes it a template of it with permissions of everyone as UserTemplateBasedVM. User B tries to create a VM based on the template that User A created. While the base VM profile can be created the storage provisioning encounters an issue. Via Template provisioning option with the thin provision option will fail due to the fact that User B does not have proper permissions to User A's storage domain. The symptom of this expected failure is the target storage domain pull-down is empty. (It really should show something or be greyed out rather than just be blank at least some sort of user notification). The real issue here is with the clone provisioning option. The idea here is to be to clone a copy of the template disks into User B's storage domain as a target where User B has poweruser role permissions. The problem here is that this fails just like the above thin provision which should not be the case. The target pulldown still blank it should by default show the target storage domain to which User B has permissions to that being Storage domain B. Further debugging yields that by assigning UserTemplateVM permissions to User A's storage domain allows User B to use either of the options above although the only one really desired is the clone option since we don't want User B creating VM's in User A's storage domain. There still however was an issue upon selecting clone and selecting Storage domain B as the target the VM is created but the disk is created in Storage domain A instead of storage domain B. Running build of the engine is built from commit: 7354d3283627bdbe30dd9c15ce45eba375280a8c - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] template provisioning permissions
Any thoughts on this one? - DHC On Wed, Mar 13, 2013 at 4:34 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Got an interesting one here as pertaining to template permissions and provisioning. Given the following setup/situation: A cluster with a user A assigned poweruser role permissions on the cluster. - User A is assigned poweruser role permissions to storage domain A - User A is a consumer of quota A which is assigned to specific storage domain A A cluster with a user B assigned poweruser role permissions on the cluster. - User B is assigned poweruser role permissions to storage domain B - User B is a consumer of quota B which is assigned to specific storage domain B User A creates a VM and makes it a template of it with permissions of everyone as UserTemplateBasedVM. User B tries to create a VM based on the template that User A created. While the base VM profile can be created the storage provisioning encounters an issue. Via Template provisioning option with the thin provision option will fail due to the fact that User B does not have proper permissions to User A's storage domain. The symptom of this expected failure is the target storage domain pull-down is empty. (It really should show something or be greyed out rather than just be blank at least some sort of user notification). The real issue here is with the clone provisioning option. The idea here is to be to clone a copy of the template disks into User B's storage domain as a target where User B has poweruser role permissions. The problem here is that this fails just like the above thin provision which should not be the case. The target pulldown still blank it should by default show the target storage domain to which User B has permissions to that being Storage domain B. Further debugging yields that by assigning UserTemplateVM permissions to User A's storage domain allows User B to use either of the options above although the only one really desired is the clone option since we don't want User B creating VM's in User A's storage domain. There still however was an issue upon selecting clone and selecting Storage domain B as the target the VM is created but the disk is created in Storage domain A instead of storage domain B. Running build of the engine is built from commit: 7354d3283627bdbe30dd9c15ce45eba375280a8c - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Userportal SPICE Woes
The hosts are setup to use secure connection so they will definitely whack out for sure. What you can do is to place the hosts into maintenance. Then make the dbase changes then restart the engine and re-add the hosts so they get configured properly with the new changes. - DHC On Wed, Mar 13, 2013 at 1:56 AM, Karli Sjöberg karli.sjob...@slu.se wrote: ** tis 2013-03-12 klockan 16:38 -0500 skrev Dead Horse: Remote console access IMHO is still a very weak point in ovirt/RHEV. SPICE is a very nice protocol but IMHO is in dire need of a platform independent client and tighter integration into the user and web portals. Something I have done to workaround issues like these is to disable spice certificate checking and SSL. psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = 'SSLEnabled' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = 'UseSecureConnectionWithServers' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name ='EnableSpiceRootCertificateValidation' - DHC I tried those modifications to the database and restarted the engine which became very upset about it and couldn´t establish connections with any Hosts. I imagine I´d have to go around all config files on all machines involved so that no one tries to use any SSL for the changes to work, Since I just wanted to use it as a temporary solution until it was fixed for real, I didn´t want to make too much of a change from the standard config, and I think that encryption is a must on a network as open as ours any way. I reverted the database changes and restarted the engine again. The only solution presented this far has been trying 3.2 instead of our current 3.1, where engine-upgrade is the road-block hindering at this point. /Karli On Fri, Mar 8, 2013 at 10:22 AM, Michal Skrivanek michal.skriva...@redhat.com wrote: On 8 Mar 2013, at 15:51, Karli Sjöberg wrote: tor 2013-03-07 klockan 13:52 +0100 skrev David Jaša: Hi, Karli Sjöberg píše v Čt 07. 03. 2013 v 10:01 +: Hi, it me again, the last time this was discussed, the problem seemed to be centred around Win/IE, but since patching to: # rpm -qa | grep ovirt ovirt-engine-config-3.1.0-4.fc17.noarch ovirt-engine-dbscripts-3.1.0-4.fc17.noarch ovirt-log-collector-3.1.0-0.git10d719.fc17.noarch ovirt-engine-restapi-3.1.0-4.fc17.noarch ovirt-image-uploader-3.1.0-0.git9c42c8.fc17.noarch ovirt-engine-genericapi-3.1.0-4.fc17.noarch ovirt-iso-uploader-3.1.0-0.git1841d9.fc17.noarch ovirt-engine-webadmin-portal-3.1.0-4.fc17.noarch ovirt-engine-setup-3.1.0-4.fc17.noarch ovirt-engine-sdk-3.2.0.2-1.fc17.noarch ovirt-engine-backend-3.1.0-4.fc17.noarch ovirt-engine-tools-common-3.1.0-4.fc17.noarch ovirt-engine-3.1.0-4.fc17.noarch ovirt-engine-userportal-3.1.0-4.fc17.noarch ovirt-engine-notification-service-3.1.0-4.fc17.noarch No one is able to connect to a spice console though the Userportal at all. Looks like no one is getting host subject - again. Tomáš? Come on guys, please tell me if there´s anything we can do to troubleshoot this issue, to help solve a real show-stopper for us! Would you be able to try it in 3.2? It was known to be broken somewhere around 3.1 release…. Thanks, michal In using oVirt, our organisation is so close to being able to ship out thin-client kiosks instead of buying expensive new HW on a yearly basis, and have our users access their desktops from practically anywhere, even from home if they´d wish. The rest of the entire system works flawlessly, just to fall flat on it´s face right on the finish line. I think that is so sad and silly, because I really love the rest of this product! We´ve set up oVirt-3.1, made a POC template, defined a pool and have been able to have them prestarted as well. But if our users aren´t able to access their console in a seamless, easy way, then what good is it? And knowing that the console works just fine from the adminportal in several different client systems (Fedora, CentOS, Win7) makes me hope that it´s just this small, yet important piece that needs to be corrected. Anxiously awaiting your response Karli Sjöberg David I´ve tried this from my workstation, running Fedora 17: $ SPICEC_DEBUG_LEVEL=0 firefox (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: could not get browser window, when trying to call OnDisconnected Here´s a collegue of mine, running Win7/IE9
[Users] template provisioning permissions
Got an interesting one here as pertaining to template permissions and provisioning. Given the following setup/situation: A cluster with a user A assigned poweruser role permissions on the cluster. - User A is assigned poweruser role permissions to storage domain A - User A is a consumer of quota A which is assigned to specific storage domain A A cluster with a user B assigned poweruser role permissions on the cluster. - User B is assigned poweruser role permissions to storage domain B - User B is a consumer of quota B which is assigned to specific storage domain B User A creates a VM and makes it a template of it with permissions of everyone as UserTemplateBasedVM. User B tries to create a VM based on the template that User A created. While the base VM profile can be created the storage provisioning encounters an issue. Via Template provisioning option with the thin provision option will fail due to the fact that User B does not have proper permissions to User A's storage domain. The symptom of this expected failure is the target storage domain pull-down is empty. (It really should show something or be greyed out rather than just be blank at least some sort of user notification). The real issue here is with the clone provisioning option. The idea here is to be to clone a copy of the template disks into User B's storage domain as a target where User B has poweruser role permissions. The problem here is that this fails just like the above thin provision which should not be the case. The target pulldown still blank it should by default show the target storage domain to which User B has permissions to that being Storage domain B. Further debugging yields that by assigning UserTemplateVM permissions to User A's storage domain allows User B to use either of the options above although the only one really desired is the clone option since we don't want User B creating VM's in User A's storage domain. There still however was an issue upon selecting clone and selecting Storage domain B as the target the VM is created but the disk is created in Storage domain A instead of storage domain B. Running build of the engine is built from commit: 7354d3283627bdbe30dd9c15ce45eba375280a8c - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Userportal SPICE Woes
Remote console access IMHO is still a very weak point in ovirt/RHEV. SPICE is a very nice protocol but IMHO is in dire need of a platform independent client and tighter integration into the user and web portals. Something I have done to workaround issues like these is to disable spice certificate checking and SSL. psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = 'SSLEnabled' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = 'UseSecureConnectionWithServers' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name ='EnableSpiceRootCertificateValidation' - DHC On Fri, Mar 8, 2013 at 10:22 AM, Michal Skrivanek michal.skriva...@redhat.com wrote: On 8 Mar 2013, at 15:51, Karli Sjöberg wrote: tor 2013-03-07 klockan 13:52 +0100 skrev David Jaša: Hi, Karli Sjöberg píše v Čt 07. 03. 2013 v 10:01 +: Hi, it me again, the last time this was discussed, the problem seemed to be centred around Win/IE, but since patching to: # rpm -qa | grep ovirt ovirt-engine-config-3.1.0-4.fc17.noarch ovirt-engine-dbscripts-3.1.0-4.fc17.noarch ovirt-log-collector-3.1.0-0.git10d719.fc17.noarch ovirt-engine-restapi-3.1.0-4.fc17.noarch ovirt-image-uploader-3.1.0-0.git9c42c8.fc17.noarch ovirt-engine-genericapi-3.1.0-4.fc17.noarch ovirt-iso-uploader-3.1.0-0.git1841d9.fc17.noarch ovirt-engine-webadmin-portal-3.1.0-4.fc17.noarch ovirt-engine-setup-3.1.0-4.fc17.noarch ovirt-engine-sdk-3.2.0.2-1.fc17.noarch ovirt-engine-backend-3.1.0-4.fc17.noarch ovirt-engine-tools-common-3.1.0-4.fc17.noarch ovirt-engine-3.1.0-4.fc17.noarch ovirt-engine-userportal-3.1.0-4.fc17.noarch ovirt-engine-notification-service-3.1.0-4.fc17.noarch No one is able to connect to a spice console though the Userportal at all. Looks like no one is getting host subject - again. Tomáš? Come on guys, please tell me if there´s anything we can do to troubleshoot this issue, to help solve a real show-stopper for us! Would you be able to try it in 3.2? It was known to be broken somewhere around 3.1 release…. Thanks, michal In using oVirt, our organisation is so close to being able to ship out thin-client kiosks instead of buying expensive new HW on a yearly basis, and have our users access their desktops from practically anywhere, even from home if they´d wish. The rest of the entire system works flawlessly, just to fall flat on it´s face right on the finish line. I think that is so sad and silly, because I really love the rest of this product! We´ve set up oVirt-3.1, made a POC template, defined a pool and have been able to have them prestarted as well. But if our users aren´t able to access their console in a seamless, easy way, then what good is it? And knowing that the console works just fine from the adminportal in several different client systems (Fedora, CentOS, Win7) makes me hope that it´s just this small, yet important piece that needs to be corrected. Anxiously awaiting your response Karli Sjöberg David I´ve tried this from my workstation, running Fedora 17: $ SPICEC_DEBUG_LEVEL=0 firefox (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: could not get browser window, when trying to call OnDisconnected Here´s a collegue of mine, running Win7/IE9: 1362648766 INFO [7196:5208] spicex_init_logger: started 1362648766 DEBUG [7196:5208] COSpiceX::put_DynamicMenu: DynamicMenu 1362648766 INFO [7196:5208] COSpiceX::put_FullScreen: New FullScreen request newVal=0x 1362648766 DEBUG [7196:5208] COSpiceX::Connect: Running spicec (C: \Users\jope\AppData\Local\virt-viewer\bin\remote-viewer.exe --spice-controller) 1362648766 INFO [7196:5208] COSpiceX::Connect: spicec pid 1640 1362648766 DEBUG [7196:5208] COSpiceX::Connect: connecting to spice client's pipe 1362648767 DEBUG [7196:5208] COSpiceX::send_init: sending init 1362648767 DEBUG [7196:5208] COSpiceX::send_wstr: sending msg id 1 : str 130.238.96.135 (1) 1362648767 DEBUG [7196:5208] COSpiceX::send_value: sending msg id 2 value 5906 1362648767 DEBUG [7196:5208] COSpiceX::send_value: sending msg id 3 value 5907 1362648767 DEBUG [7196:5208] COSpiceX::send_wstr: sending password 1362648767 DEBUG [7196:5208] COSpiceX::send_wstr: sending msg id 11 : str IT-Avdelningen-1:%d - Press SHIFT+F12 to Release Cursor (1) 1362648767 DEBUG [7196:5208] COSpiceX::send_value: sending msg id 10 value 1 1362648767 DEBUG
Re: [Users] Userportal SPICE Woes
Itamar, Nail on the head. Pretty much what I describe here is sorely needed: http://lists.ovirt.org/pipermail/users/2013-February/012562.html - DHC On Tue, Mar 12, 2013 at 5:53 PM, Itamar Heim ih...@redhat.com wrote: On 03/12/2013 11:38 PM, Dead Horse wrote: Remote console access IMHO is still a very weak point in ovirt/RHEV. SPICE is a very nice protocol but IMHO is in dire need of a platform independent client and tighter integration into the user and web tighter integration or support non integrated clients (the mime based approach which won't require the XPI/ActiveX)? portals. Something I have done to workaround issues like these is to disable spice certificate checking and SSL. psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = 'SSLEnabled' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name = '**UseSecureConnectionWithServers**' psql engine -U postgres -c UPDATE vdc_options set option_value = 'false' where option_name ='**EnableSpiceRootCertificateVali**dation' - DHC On Fri, Mar 8, 2013 at 10:22 AM, Michal Skrivanek michal.skriva...@redhat.com mailto:michal.skrivanek@**redhat.commichal.skriva...@redhat.com wrote: On 8 Mar 2013, at 15:51, Karli Sjöberg wrote: tor 2013-03-07 klockan 13:52 +0100 skrev David Jaša: Hi, Karli Sjöberg píše v Čt 07. 03. 2013 v 10:01 +: Hi, it me again, the last time this was discussed, the problem seemed to be centred around Win/IE, but since patching to: # rpm -qa | grep ovirt ovirt-engine-config-3.1.0-4.**fc17.noarch ovirt-engine-dbscripts-3.1.0-**4.fc17.noarch ovirt-log-collector-3.1.0-0.**git10d719.fc17.noarch ovirt-engine-restapi-3.1.0-4.**fc17.noarch ovirt-image-uploader-3.1.0-0.**git9c42c8.fc17.noarch ovirt-engine-genericapi-3.1.0-**4.fc17.noarch ovirt-iso-uploader-3.1.0-0.**git1841d9.fc17.noarch ovirt-engine-webadmin-portal-**3.1.0-4.fc17.noarch ovirt-engine-setup-3.1.0-4.**fc17.noarch ovirt-engine-sdk-3.2.0.2-1.**fc17.noarch ovirt-engine-backend-3.1.0-4.**fc17.noarch ovirt-engine-tools-common-3.1.**0-4.fc17.noarch ovirt-engine-3.1.0-4.fc17.**noarch ovirt-engine-userportal-3.1.0-**4.fc17.noarch ovirt-engine-notification-**service-3.1.0-4.fc17.noarch No one is able to connect to a spice console though the Userportal at all. Looks like no one is getting host subject - again. Tomáš? Come on guys, please tell me if there´s anything we can do to troubleshoot this issue, to help solve a real show-stopper for us! Would you be able to try it in 3.2? It was known to be broken somewhere around 3.1 release…. Thanks, michal In using oVirt, our organisation is so close to being able to ship out thin-client kiosks instead of buying expensive new HW on a yearly basis, and have our users access their desktops from practically anywhere, even from home if they´d wish. The rest of the entire system works flawlessly, just to fall flat on it´s face right on the finish line. I think that is so sad and silly, because I really love the rest of this product! We´ve set up oVirt-3.1, made a POC template, defined a pool and have been able to have them prestarted as well. But if our users aren´t able to access their console in a seamless, easy way, then what good is it? And knowing that the console works just fine from the adminportal in several different client systems (Fedora, CentOS, Win7) makes me hope that it´s just this small, yet important piece that needs to be corrected. Anxiously awaiting your response Karli Sjöberg David I´ve tried this from my workstation, running Fedora 17: ==**==** $ SPICEC_DEBUG_LEVEL=0 firefox (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: controller connect: Filen eller katalogen finns inte (plugin-container:18293): SpiceXPI-CRITICAL **: could not get browser window, when trying to call OnDisconnected ==**==** Here´s a collegue of mine, running Win7/IE9: ==**==** 1362648766 INFO [7196:5208] spicex_init_logger: started 1362648766 DEBUG [7196:5208] COSpiceX::put_DynamicMenu: DynamicMenu 1362648766 INFO [7196:5208] COSpiceX::put_FullScreen: New FullScreen request newVal=0x 1362648766 DEBUG [7196:5208] COSpiceX::Connect: Running spicec (C: \Users\jope\AppData\Local\**virt-viewer\bin\remote-viewer
Re: [Users] VNC web console
My 2 cents here. I think what he masses would like to ultimately see here is a pure web based SPICE/VNC client nicely integrated into the user/admin portals. In this regard the console becomes truly portable across operating systems and browsers. The biggest win here is no longer having to be stuck with Internet Exploder as the only browser on a windows client that can access a spice console. At least on a Linux based client we have Firefox, but chrome would be nice. Due to the current console situation what is completely out in the cold here is tablets (granted you can use a vnc client on a tablet, but painful due to manual connection steps ip/port/password ATM). Likewise VNC != SPICE quality and interactivity wise on top of the other obvious advantages or SPICE over VNC (EG: HW Accel via QXL). Also left out here are OSX users for which a solution like this would console enable them as well. - DHC On Thu, Feb 21, 2013 at 7:16 AM, Michal Skrivanek michal.skriva...@redhat.com wrote: On Feb 18, 2013, at 10:10 , Jaco ubuntumu...@gmail.com wrote: I've seen this in the list archives, but not sure how much traction it's gained in the intervening period (or ). Hi, still pending, but we're getting there:-) I've fiddled with, and had some pretty decent experiences with Guacamole: http://guac-dev.org/ An HTML(5?)-based VNC client for a browser - no client-side components needed :) how would you compare it to novnc? RDP support might be handy…on the other hand at first sight looking at http://sourceforge.net/projects/guacamole/forums/forum/1110834/topic/5236646and http://cloudiad.com/analysis/doc1 it looks noVNC would be easier to integrate with. Especially when qemu just recently got direct support for websockets. Then we really do not need anything at the engine side at all. Thanks, michal Could fit in nicely with the project. Just putting it out there - J ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Power Users cannot add Network Interfaces
I will spool up a build this morning and verify. - DHC On Tue, Feb 19, 2013 at 8:36 AM, Alona Kaplan alkap...@redhat.com wrote: Hi DHC, Thank you for reporting the bug. The bug was fixed in commit 236f90f44add695def6932c609fd9935524c15ae http://gerrit.ovirt.org/#/c/12137/ Alona. - Original Message - From: Moti Asayag masa...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users users@ovirt.org Sent: Sunday, February 17, 2013 9:56:11 AM Subject: Re: [Users] Power Users cannot add Network Interfaces On 02/15/2013 11:02 PM, Dead Horse wrote: Did a fresh engine install/setup of a newer build from master and this behavior is still persists. Fresh install/setup using the older build from the commit I mentioned previously I can confirm that the the issue is not present, things work as expected. - DHC This is a bug in User Portal. I've filed a bug for it: https://bugzilla.redhat.com/show_bug.cgi?id=912005 I guess it will be fixed in a short while. On Thu, Feb 14, 2013 at 3:58 PM, Dead Horse deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com wrote: It is still marked and recognized as Management Network and a VM Network. SuperUsers can still add VNIC's this is only affecting non-superusers EG: PowerUserRole - DHC On Thu, Feb 14, 2013 at 3:30 PM, Moti Asayag masa...@redhat.com mailto:masa...@redhat.com wrote: On 02/14/2013 06:13 PM, Dead Horse wrote: The dialog is graying the ok button, screenshot attached. Also attached is the engine log, although I don't see anything in it that jumps out at me. - DHC Is your management network a non-vm network ? The described symptom looks a like [1]. [1] https://bugzilla.redhat.com/show_bug.cgi?id=95 On Thu, Feb 14, 2013 at 3:35 AM, Moti Asayag masa...@redhat.com mailto:masa...@redhat.com mailto:masa...@redhat.com mailto:masa...@redhat.com wrote: On 02/14/2013 12:47 AM, Eli Mesika wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com To: users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org Sent: Thursday, February 14, 2013 12:16:58 AM Subject: [Users] Power Users cannot add Network Interfaces Power Users seems to no longer be able to add Network Interfaces to Virtual Machines. Super Users of course have no issues. The power users have been granted NetworkUser Role on the respective Networks, PowerUserRole on the respective Clusters/Storage Domains. This failure mode is with the currently running ovirt-engine I built from commit: 899d96c055784a54b6d2b8181e5ed13b258d009f This used to work with my previous engine build from commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b Hi Can you please attach vdsm/engine log so we will be able to track the error thrown and check which commit caused that ? I guess the error would be a can-do-action message: The user is not authorized to perform this action. I'll try reproducing it on my environment to see what went wrong. - DHC ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org mailto:Users@ovirt.org mailto:Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org mailto:Users
Re: [Users] Fwd: Internet Explorer with latest master
Rebuilding with -Dgwt.userAgent= ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep had no effect. Which given that the default behavior of master is to do this is not surprising. The Internet Exploder edge mode seems to be the core culprit here. - DHC On Fri, Feb 15, 2013 at 12:18 PM, Daniel Erez de...@redhat.com wrote: - Original Message - From: Vojtech Szocs vsz...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users users@ovirt.org Sent: Friday, February 15, 2013 3:32:03 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master Hi, doing make rpm will actually build Engine using Maven and produce RPMs, but you might also want to deploy Engine EAR that was built to JBoss server via make install_artifacts. Thus I assume the master build does not default to compiling for all browsers? Aside from make rpm-quick, which builds quick dirty RPMs for Firefox [gwt.userAgent=gecko1_8], I don't see any gwt.userAgent stuff in makefile, so the master build defaults to compiling for all browsers, unless you override this via EXTRA_BUILD_FLAGS. I think an easier way to do this is as Alona suggested, invoking Maven build directly: $ mvn clean install -Pdep,gwt-admin,gwt-user, but it's up to you. Anyway, I think the problem isn't related to gwt.userAgent property at all. If the GWT permutation for IE is missing, blank page will be shown, as GWT permutation selector script (*.nocache.js) will not be able to load that permutation. I think the problem might be caused by your client's IE browser settings. As per [1], IE should render HTML page with !DOCTYPE html using standards mode; rendering mode which WebAdmin and UserPortal require. So if you see garbled mess in IE, it means IE doesn't render HTML page in standards mode, but in other (quirks?) mode, due to some IE browser settings override. The X-UA-Compatible meta-tag actually *forces* IE to render HTML page in standards mode, which overrides any IE browser settings: meta http-equiv=X-UA-Compatible content=IE=9/ (I think we should reconsider adding this meta-tag due to reasons above, to ensure better compatibility with IE.) Instead of forcing IE9 mode we can use edge mode to display content in the highest mode available. I.e.: meta http-equiv=X-UA-Compatible content=IE=edge/ Vojtech [1] http://msdn.microsoft.com/en-us/library/cc288325(v=vs.85).aspx - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alona Kaplan alkap...@redhat.com Cc: users users@ovirt.org, Vojtech Szocs vsz...@redhat.com Sent: Thursday, February 14, 2013 8:02:04 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master NVM got it -- BUILD_FLAGS=-Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Building now. - DHC On Thu, Feb 14, 2013 at 12:23 PM, Dead Horse deadhorseconsult...@gmail.com wrote: I remember there being a GWT configuration file somewhere where those could be set? - DHC On Thu, Feb 14, 2013 at 10:44 AM, Alona Kaplan alkap...@redhat.com wrote: Hi, Instead of doing make rpm please compile manually using the following command: mvn clean install -Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Vojtech Szocs vsz...@redhat.com Cc: Alona Kaplan alkap...@redhat.com , users@ovirt.org users@ovirt.org Sent: Thursday, February 14, 2013 6:18:54 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master When I compile the engine I do the following: (FC17 build host) -- git clone http://gerrit.ovirt.org/ovirt-engine.git -- cd ovirt-engine -- make rpm Thus I assume the master build does not default to compiling for all browsers? The GWT compliation seems to go through quite a few permuatations during the compile. If master is not set to compile for all browsers then given the above build steps what additionally should I be doing to make it so (guessing refer to above but how to simplify to make rpm) ? - DHC On Thu, Feb 14, 2013 at 10:09 AM, Vojtech Szocs vsz...@redhat.com wrote: Hi, Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? this is a good point, please make sure to compile UserPortal for all browsers, e.g. instead of: $ mvn clean install -Pdep,gwt-user -Dgwt.userAgent=specificBrowser you should just do: $ mvn clean install -Pdep,gwt-user This will compile UserPortal for all browsers. However, I don't think this is the problem here, because IIRC when you compile UserPortal only for browser X and you try to access it in browser Y
Re: [Users] Fwd: Internet Explorer with latest master
Disabling compatibility view does the trick. - DHC On Fri, Feb 15, 2013 at 2:06 PM, Vojtech Szocs vsz...@redhat.com wrote: Hi, The Internet Exploder edge mode seems to be the core culprit here. please check http://stackoverflow.com/a/5673902, intranet sites in IE8/IE9 are *always* rendered in compatibility (non-standards) mode. This might be the problem you're facing with WebAdmin and UserPortal. You might want to check IE Compatibility View Settings dialog box and ensure that nothing gets displayed in compatibility view. Can you please try: - check IE Compatibility View Settings and see if it works, without any X-UA-Compatible stuff whatsoever - if not, adding IE=edge and see if it works Vojtech - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Daniel Erez de...@redhat.com Cc: Vojtech Szocs vsz...@redhat.com, users users@ovirt.org Sent: Friday, February 15, 2013 8:45:09 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master Rebuilding with -Dgwt.userAgent= ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep had no effect. Which given that the default behavior of master is to do this is not surprising. The Internet Exploder edge mode seems to be the core culprit here. - DHC On Fri, Feb 15, 2013 at 12:18 PM, Daniel Erez de...@redhat.com wrote: - Original Message - From: Vojtech Szocs vsz...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users users@ovirt.org Sent: Friday, February 15, 2013 3:32:03 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master Hi, doing make rpm will actually build Engine using Maven and produce RPMs, but you might also want to deploy Engine EAR that was built to JBoss server via make install_artifacts. Thus I assume the master build does not default to compiling for all browsers? Aside from make rpm-quick, which builds quick dirty RPMs for Firefox [gwt.userAgent=gecko1_8], I don't see any gwt.userAgent stuff in makefile, so the master build defaults to compiling for all browsers, unless you override this via EXTRA_BUILD_FLAGS. I think an easier way to do this is as Alona suggested, invoking Maven build directly: $ mvn clean install -Pdep,gwt-admin,gwt-user, but it's up to you. Anyway, I think the problem isn't related to gwt.userAgent property at all. If the GWT permutation for IE is missing, blank page will be shown, as GWT permutation selector script (*.nocache.js) will not be able to load that permutation. I think the problem might be caused by your client's IE browser settings. As per [1], IE should render HTML page with !DOCTYPE html using standards mode; rendering mode which WebAdmin and UserPortal require. So if you see garbled mess in IE, it means IE doesn't render HTML page in standards mode, but in other (quirks?) mode, due to some IE browser settings override. The X-UA-Compatible meta-tag actually *forces* IE to render HTML page in standards mode, which overrides any IE browser settings: meta http-equiv=X-UA-Compatible content=IE=9/ (I think we should reconsider adding this meta-tag due to reasons above, to ensure better compatibility with IE.) Instead of forcing IE9 mode we can use edge mode to display content in the highest mode available. I.e.: meta http-equiv=X-UA-Compatible content=IE=edge/ Vojtech [1] http://msdn.microsoft.com/en-us/library/cc288325(v=vs.85).aspx - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alona Kaplan alkap...@redhat.com Cc: users users@ovirt.org , Vojtech Szocs vsz...@redhat.com Sent: Thursday, February 14, 2013 8:02:04 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master NVM got it -- BUILD_FLAGS=-Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Building now. - DHC On Thu, Feb 14, 2013 at 12:23 PM, Dead Horse deadhorseconsult...@gmail.com wrote: I remember there being a GWT configuration file somewhere where those could be set? - DHC On Thu, Feb 14, 2013 at 10:44 AM, Alona Kaplan alkap...@redhat.com wrote: Hi, Instead of doing make rpm please compile manually using the following command: mvn clean install -Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Vojtech Szocs vsz...@redhat.com Cc: Alona Kaplan alkap...@redhat.com , users@ovirt.org users@ovirt.org Sent: Thursday, February 14, 2013 6:18:54 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master When I compile the engine I do the following: (FC17 build host) -- git clone http://gerrit.ovirt.org/ovirt-engine.git -- cd ovirt-engine -- make rpm Thus I assume the master build does
Re: [Users] Power Users cannot add Network Interfaces
Did a fresh engine install/setup of a newer build from master and this behavior is still persists. Fresh install/setup using the older build from the commit I mentioned previously I can confirm that the the issue is not present, things work as expected. - DHC On Thu, Feb 14, 2013 at 3:58 PM, Dead Horse deadhorseconsult...@gmail.comwrote: It is still marked and recognized as Management Network and a VM Network. SuperUsers can still add VNIC's this is only affecting non-superusers EG: PowerUserRole - DHC On Thu, Feb 14, 2013 at 3:30 PM, Moti Asayag masa...@redhat.com wrote: On 02/14/2013 06:13 PM, Dead Horse wrote: The dialog is graying the ok button, screenshot attached. Also attached is the engine log, although I don't see anything in it that jumps out at me. - DHC Is your management network a non-vm network ? The described symptom looks a like [1]. [1] https://bugzilla.redhat.com/show_bug.cgi?id=95 On Thu, Feb 14, 2013 at 3:35 AM, Moti Asayag masa...@redhat.com mailto:masa...@redhat.com wrote: On 02/14/2013 12:47 AM, Eli Mesika wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com To: users@ovirt.org mailto:users@ovirt.org users@ovirt.org mailto:users@ovirt.org Sent: Thursday, February 14, 2013 12:16:58 AM Subject: [Users] Power Users cannot add Network Interfaces Power Users seems to no longer be able to add Network Interfaces to Virtual Machines. Super Users of course have no issues. The power users have been granted NetworkUser Role on the respective Networks, PowerUserRole on the respective Clusters/Storage Domains. This failure mode is with the currently running ovirt-engine I built from commit: 899d96c055784a54b6d2b8181e5ed13b258d009f This used to work with my previous engine build from commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b Hi Can you please attach vdsm/engine log so we will be able to track the error thrown and check which commit caused that ? I guess the error would be a can-do-action message: The user is not authorized to perform this action. I'll try reproducing it on my environment to see what went wrong. - DHC ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Fwd: Internet Explorer with latest master
When I compile the engine I do the following: (FC17 build host) -- git clone http://gerrit.ovirt.org/ovirt-engine.git -- cd ovirt-engine -- make rpm Thus I assume the master build does not default to compiling for all browsers? The GWT compliation seems to go through quite a few permuatations during the compile. If master is not set to compile for all browsers then given the above build steps what additionally should I be doing to make it so (guessing refer to above but how to simplify to make rpm) ? - DHC On Thu, Feb 14, 2013 at 10:09 AM, Vojtech Szocs vsz...@redhat.com wrote: Hi, Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? this is a good point, please make sure to compile UserPortal for all browsers, e.g. instead of: $ mvn clean install -Pdep,gwt-user -Dgwt.userAgent=specificBrowser you should just do: $ mvn clean install -Pdep,gwt-user This will compile UserPortal for all browsers. However, I don't think this is the problem here, because IIRC when you compile UserPortal only for browser X and you try to access it in browser Y, you'll get an empty webpage. Vojtech - Original Message - From: Alona Kaplan alkap...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users@ovirt.org Sent: Thursday, February 14, 2013 4:18:36 PM Subject: [Users] Fwd: Internet Explorer with latest master Hi, The webadmin and userportal works pretty fine for me on ie8 and ie9. Can you please be more specific about what doesn't work for you? Maybe attaching a screen shot. Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? You should compile without any specific gwt.userAgent to support all the browsers or to add -Dgwt.userAgent=ie8,ie9 to support ie8 and ie9. Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Wednesday, February 13, 2013 7:01:48 PM Subject: [Users] Internet Explorer with latest master Ever since this change: http://gerrit.ovirt.org/#/c/11078/ The webadmin and userportal UI no longer displays nor works correctly within Internet Explorer 8 or 9. Under IE 8 or 9 the UI is pretty much an usable garbled mess. Everything looks and works fine under Chrome and Firefox. Unfortunately Internet Exploder is a necessary evil for windows users to be able to get at SPICE consoles. (No spice plugin/support for Firefox/Chrome yet...) - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Fwd: Internet Explorer with latest master
I remember there being a GWT configuration file somewhere where those could be set? - DHC On Thu, Feb 14, 2013 at 10:44 AM, Alona Kaplan alkap...@redhat.com wrote: Hi, Instead of doing make rpm please compile manually using the following command: mvn clean install -Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Vojtech Szocs vsz...@redhat.com Cc: Alona Kaplan alkap...@redhat.com, users@ovirt.org users@ovirt.org Sent: Thursday, February 14, 2013 6:18:54 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master When I compile the engine I do the following: (FC17 build host) -- git clone http://gerrit.ovirt.org/ovirt-engine.git -- cd ovirt-engine -- make rpm Thus I assume the master build does not default to compiling for all browsers? The GWT compliation seems to go through quite a few permuatations during the compile. If master is not set to compile for all browsers then given the above build steps what additionally should I be doing to make it so (guessing refer to above but how to simplify to make rpm) ? - DHC On Thu, Feb 14, 2013 at 10:09 AM, Vojtech Szocs vsz...@redhat.com wrote: Hi, Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? this is a good point, please make sure to compile UserPortal for all browsers, e.g. instead of: $ mvn clean install -Pdep,gwt-user -Dgwt.userAgent=specificBrowser you should just do: $ mvn clean install -Pdep,gwt-user This will compile UserPortal for all browsers. However, I don't think this is the problem here, because IIRC when you compile UserPortal only for browser X and you try to access it in browser Y, you'll get an empty webpage. Vojtech - Original Message - From: Alona Kaplan alkap...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users@ovirt.org Sent: Thursday, February 14, 2013 4:18:36 PM Subject: [Users] Fwd: Internet Explorer with latest master Hi, The webadmin and userportal works pretty fine for me on ie8 and ie9. Can you please be more specific about what doesn't work for you? Maybe attaching a screen shot. Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? You should compile without any specific gwt.userAgent to support all the browsers or to add -Dgwt.userAgent=ie8,ie9 to support ie8 and ie9. Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Wednesday, February 13, 2013 7:01:48 PM Subject: [Users] Internet Explorer with latest master Ever since this change: http://gerrit.ovirt.org/#/c/11078/ The webadmin and userportal UI no longer displays nor works correctly within Internet Explorer 8 or 9. Under IE 8 or 9 the UI is pretty much an usable garbled mess. Everything looks and works fine under Chrome and Firefox. Unfortunately Internet Exploder is a necessary evil for windows users to be able to get at SPICE consoles. (No spice plugin/support for Firefox/Chrome yet...) - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Fwd: Internet Explorer with latest master
NVM got it -- BUILD_FLAGS=-Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Building now. - DHC On Thu, Feb 14, 2013 at 12:23 PM, Dead Horse deadhorseconsult...@gmail.comwrote: I remember there being a GWT configuration file somewhere where those could be set? - DHC On Thu, Feb 14, 2013 at 10:44 AM, Alona Kaplan alkap...@redhat.comwrote: Hi, Instead of doing make rpm please compile manually using the following command: mvn clean install -Dgwt.userAgent=ie8,ie9,gecko1_8,safari,opera -Pgwt-admin,gwt-user,dep Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Vojtech Szocs vsz...@redhat.com Cc: Alona Kaplan alkap...@redhat.com, users@ovirt.org users@ovirt.org Sent: Thursday, February 14, 2013 6:18:54 PM Subject: Re: [Users] Fwd: Internet Explorer with latest master When I compile the engine I do the following: (FC17 build host) -- git clone http://gerrit.ovirt.org/ovirt-engine.git -- cd ovirt-engine -- make rpm Thus I assume the master build does not default to compiling for all browsers? The GWT compliation seems to go through quite a few permuatations during the compile. If master is not set to compile for all browsers then given the above build steps what additionally should I be doing to make it so (guessing refer to above but how to simplify to make rpm) ? - DHC On Thu, Feb 14, 2013 at 10:09 AM, Vojtech Szocs vsz...@redhat.com wrote: Hi, Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? this is a good point, please make sure to compile UserPortal for all browsers, e.g. instead of: $ mvn clean install -Pdep,gwt-user -Dgwt.userAgent=specificBrowser you should just do: $ mvn clean install -Pdep,gwt-user This will compile UserPortal for all browsers. However, I don't think this is the problem here, because IIRC when you compile UserPortal only for browser X and you try to access it in browser Y, you'll get an empty webpage. Vojtech - Original Message - From: Alona Kaplan alkap...@redhat.com To: Dead Horse deadhorseconsult...@gmail.com Cc: users@ovirt.org Sent: Thursday, February 14, 2013 4:18:36 PM Subject: [Users] Fwd: Internet Explorer with latest master Hi, The webadmin and userportal works pretty fine for me on ie8 and ie9. Can you please be more specific about what doesn't work for you? Maybe attaching a screen shot. Maybe you've compiled the application with -Dgwt.userAgent flag for a specific browser? You should compile without any specific gwt.userAgent to support all the browsers or to add -Dgwt.userAgent=ie8,ie9 to support ie8 and ie9. Alona. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Wednesday, February 13, 2013 7:01:48 PM Subject: [Users] Internet Explorer with latest master Ever since this change: http://gerrit.ovirt.org/#/c/11078/ The webadmin and userportal UI no longer displays nor works correctly within Internet Explorer 8 or 9. Under IE 8 or 9 the UI is pretty much an usable garbled mess. Everything looks and works fine under Chrome and Firefox. Unfortunately Internet Exploder is a necessary evil for windows users to be able to get at SPICE consoles. (No spice plugin/support for Firefox/Chrome yet...) - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Power Users cannot add Network Interfaces
It is still marked and recognized as Management Network and a VM Network. SuperUsers can still add VNIC's this is only affecting non-superusers EG: PowerUserRole - DHC On Thu, Feb 14, 2013 at 3:30 PM, Moti Asayag masa...@redhat.com wrote: On 02/14/2013 06:13 PM, Dead Horse wrote: The dialog is graying the ok button, screenshot attached. Also attached is the engine log, although I don't see anything in it that jumps out at me. - DHC Is your management network a non-vm network ? The described symptom looks a like [1]. [1] https://bugzilla.redhat.com/show_bug.cgi?id=95 On Thu, Feb 14, 2013 at 3:35 AM, Moti Asayag masa...@redhat.com mailto:masa...@redhat.com wrote: On 02/14/2013 12:47 AM, Eli Mesika wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com mailto:deadhorseconsult...@gmail.com To: users@ovirt.org mailto:users@ovirt.org users@ovirt.org mailto:users@ovirt.org Sent: Thursday, February 14, 2013 12:16:58 AM Subject: [Users] Power Users cannot add Network Interfaces Power Users seems to no longer be able to add Network Interfaces to Virtual Machines. Super Users of course have no issues. The power users have been granted NetworkUser Role on the respective Networks, PowerUserRole on the respective Clusters/Storage Domains. This failure mode is with the currently running ovirt-engine I built from commit: 899d96c055784a54b6d2b8181e5ed13b258d009f This used to work with my previous engine build from commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b Hi Can you please attach vdsm/engine log so we will be able to track the error thrown and check which commit caused that ? I guess the error would be a can-do-action message: The user is not authorized to perform this action. I'll try reproducing it on my environment to see what went wrong. - DHC ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org mailto:Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Internet Explorer with latest master
Ever since this change: http://gerrit.ovirt.org/#/c/11078/ The webadmin and userportal UI no longer displays nor works correctly within Internet Explorer 8 or 9. Under IE 8 or 9 the UI is pretty much an usable garbled mess. Everything looks and works fine under Chrome and Firefox. Unfortunately Internet Exploder is a necessary evil for windows users to be able to get at SPICE consoles. (No spice plugin/support for Firefox/Chrome yet...) - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
Current master does resolve the issue. However in order to test it the zombie reaper patch from the ovirt-3.2 branch must be applied to avoid that issue. - DHC On Fri, Feb 8, 2013 at 2:08 AM, Vinzenz Feenstra vfeen...@redhat.comwrote: On 02/05/2013 10:57 PM, Dead Horse wrote: Confirmed ovirt-3.2 branch of vdsm does work with migrations. So there is a difference between it and master as pertains to the migration logic. - DHC Hi, could you please retry it with the current master? We have fixed the issue which caused this regression. :-) Thank you :) On Tue, Feb 5, 2013 at 2:26 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Dan, Building and testing it now. - DHC On Tue, Feb 5, 2013 at 2:39 AM, Dan Kenigsberg dan...@redhat.com wrote: On Mon, Feb 04, 2013 at 10:38:16AM -0600, Dead Horse wrote: VDSM built from commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4 works Current VDSM built from master does not work. I could try spending some time trying to bisect and find out where the breakage occurred I suppose. Would you be kind to find the time to help us here? Clearly, the commit on top of c343e1833f7b6e5428dd90f14f7807dca1baa0b4 introduces nasty supervdsm regressions; it has to be reverted for any meaningful testing. However I do not see how it can be related to the problem at hand. Would you at least try out the ovirt-3.2 branch (where the infamous zombie reaper commit is reverted)? Dan. ___ Users mailing listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users -- Regards, Vinzenz Feenstra | Senior Software Engineer RedHat Engineering Virtualization R D Phone: +420 532 294 625 IRC: vfeenstr or evilissimo Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
Dan, Building and testing it now. - DHC On Tue, Feb 5, 2013 at 2:39 AM, Dan Kenigsberg dan...@redhat.com wrote: On Mon, Feb 04, 2013 at 10:38:16AM -0600, Dead Horse wrote: VDSM built from commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4 works Current VDSM built from master does not work. I could try spending some time trying to bisect and find out where the breakage occurred I suppose. Would you be kind to find the time to help us here? Clearly, the commit on top of c343e1833f7b6e5428dd90f14f7807dca1baa0b4 introduces nasty supervdsm regressions; it has to be reverted for any meaningful testing. However I do not see how it can be related to the problem at hand. Would you at least try out the ovirt-3.2 branch (where the infamous zombie reaper commit is reverted)? Dan. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
Confirmed ovirt-3.2 branch of vdsm does work with migrations. So there is a difference between it and master as pertains to the migration logic. - DHC On Tue, Feb 5, 2013 at 2:26 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Dan, Building and testing it now. - DHC On Tue, Feb 5, 2013 at 2:39 AM, Dan Kenigsberg dan...@redhat.com wrote: On Mon, Feb 04, 2013 at 10:38:16AM -0600, Dead Horse wrote: VDSM built from commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4 works Current VDSM built from master does not work. I could try spending some time trying to bisect and find out where the breakage occurred I suppose. Would you be kind to find the time to help us here? Clearly, the commit on top of c343e1833f7b6e5428dd90f14f7807dca1baa0b4 introduces nasty supervdsm regressions; it has to be reverted for any meaningful testing. However I do not see how it can be related to the problem at hand. Would you at least try out the ovirt-3.2 branch (where the infamous zombie reaper commit is reverted)? Dan. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
VDSM built from commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4 works Current VDSM built from master does not work. I could try spending some time trying to bisect and find out where the breakage occurred I suppose. - DHC On Sun, Feb 3, 2013 at 10:13 AM, Martin Kletzander mklet...@redhat.comwrote: On 02/03/2013 08:40 AM, Dan Kenigsberg wrote: On Fri, Feb 01, 2013 at 11:44:08PM +0100, Martin Kletzander wrote: On 02/01/2013 09:29 PM, Dead Horse wrote: To test further I loaded up two more identical servers with EL 6.3 and the same package versions originally indicated. The difference here is that I did not turn these into ovirt nodes. EG: installing VDSM. - All configurations were left at defaults on both servers - iptables and selinux disabled on both servers - verified full connectivty between both servers - setup ssh (/root/authorized keys) between the servers -- this turned out to be the key! Then using syntax found here: http://libvirt.org/migration.html#flowpeer2peer EG: From the source server I issued the following: So your client equals to the source server, that makes us sure that the connection is made on the same network for p2p and non-p2p migration. virsh migrate --p2p sl63 qemu+ssh://192.168.1.2/system You're using ssh transport here, but isn't vdsm using tcp or tls? It is! So then testing it with '+ssh' does not help much. But at least we know the addresses are reachable. According to the config file tcp transport is enabled with no authentication whatsoever... It fails in exactly the same way as previously indicated when the destination server does not have an ssh rsa pub ID from the source system in it's /root/.ssh/authorized_keys file. However once the ssh rsa pub ID is in place on the destination system all is well and migrations work as expected. ..., which would mean you need no ssh keys when migrating using tcp transport instead. Also during p2p migration the source libvirt daemon can't ask you for the password, but when not using p2p the client is connecting to the destination, thus being able to ask for the password and/or use different ssh keys. But it looks like none of this has anything to do with the problem as: 1) as you found out, changing vdsm versions makes the problem go away/appear and I've missed this point. Which version of vdsm makes it go away? Sorry, I've got it stuck in my head that part of the thread was about it, but when going through the mail now it makes less sense than before. I probably understood that from [1] and maybe some other sentence that mixed in my head, but was related to the ssh migration. Sorry for that, Martin [1] http://www.mail-archive.com/users@ovirt.org/msg06105.html ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
Both nodes are identical and can fully communicate with each other. Since the normal non p2p live migration works both hosts can reach each other via the connection URI. Perhaps I am missing something here? - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
Doh, brain fart VDSM is not involved here for the purposed of the needed test. Here is my initial whack at it: Source Node: virsh # list IdName State 1 sl63 running virsh # migrate --p2p sl63 qemu+ssh://192.168.1.2/system error: operation failed: Failed to connect to remote libvirt URI qemu+ssh:// 3.57.111.32/system virsh # migrate --live sl63 qemu+ssh://192.168.1.2/system The authenticity of host '3.57.111.32 (192.168.1.2)' can't be established. RSA key fingerprint is e5:1d:b3:e5:38:5f:e1:8b:73:26:9e:15:c8:0a:2d:ac. Are you sure you want to continue connecting (yes/no)? yes root@192.168.1.2's password: Please enter your authentication name: vdsm@ovirt Please enter your password: virsh # Dest Node After migrate --live: virsh # list IdName State 2 sl63 running virsh # On Thu, Jan 31, 2013 at 10:38 AM, Dead Horse deadhorseconsult...@gmail.comwrote: Shu, I build oVirt Engine and vdsm from source myself. The commits I indicated are what I built from.I run the engine under FC17 and my nodes are running EL6.x respectively. Dan, I reverted VDSM on my two test nodes to an earlier build of VDSM (commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4) VDSM after the above commit is broken due to commit: fc3a44f71d2ef202cff18d7203b9e4165b546621 however when I built and tested from master yesterday I did apply a patch I tested for ybronhei which fixed that issue. I will build VDSM from master, today w/ the supervdsm patch and try the manual migration you indicated. - DHC On Thu, Jan 31, 2013 at 4:56 AM, Dan Kenigsberg dan...@redhat.com wrote: On Thu, Jan 31, 2013 at 11:08:58AM +0100, Martin Kletzander wrote: On 01/31/2013 10:25 AM, Dan Kenigsberg wrote: On Thu, Jan 31, 2013 at 09:43:44AM +0100, Martin Kletzander wrote: On 01/30/2013 08:40 PM, Dead Horse wrote: The nodes are EL6.3 based. Currently installed libvirt packages: libvirt-lock-sanlock-0.9.10-21.el6_3.8.x86_64 libvirt-cim-0.6.1-3.el6.x86_64 libvirt-0.9.10-21.el6_3.8.x86_64 libvirt-python-0.9.10-21.el6_3.8.x86_64 libvirt-client-0.9.10-21.el6_3.8.x86_64 and qemu packages: qemu-kvm-0.12.1.2-2.295.el6_3.10.x86_64 qemu-kvm-tools-0.12.1.2-2.295.el6_3.10.x86_64 qemu-img-0.12.1.2-2.295.el6_3.10.x86_64 Thus my presumption here given the above is that virDomainMigrateToURI2 has not yet been patched and/or back-ported into the EL6.x libvirt/qemu? virDomainMigrateToURI2 is supported since 0.9.2, but is there a possibility the code is requesting direct migration? That might explain the message, which is then incorrect; this was fixed in [1]. Martin [1] http://libvirt.org/git/?p=libvirt.git;a=commitdiff;h=3189dfb1636da22d426d2fc07cc9f60304b16c5c What is direct migration exactly, in the context of qemu-kvm? We are using p2p migration http://gerrit.ovirt.org/gitweb?p=vdsm.git;a=blob;f=vdsm/libvirtvm.py;h=fe140ecbfac665248e2ad5c4bfaebaf54ab884cc;hb=18c24f7c7c27ac732c4a760caa9524e7319cd47e#l501 OK, so that's not the issue, sorry for the confusion. I was thinking it would somehow get there. Direct migration doesn't exist in QEMU at all, so it seemed weird, but I can't seem to find any other reason for this failure; will keep searching, though. In this case, Dead Horse, would you try to migrate a VM (that you do not care much about) using virsh -c qemu+tls://hostname/system migrate --p2p dsthost? I'd like to see that the problem reproduces this way, too. More of libvirtd.log may help. You may want to disable iptables for a moment, just to eliminate a common cause of failure. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
note ignore the IP diff in the ssh host auth -- copy/paste fail ;) - DHC On Thu, Jan 31, 2013 at 11:25 AM, Dead Horse deadhorseconsult...@gmail.comwrote: Doh, brain fart VDSM is not involved here for the purposed of the needed test. Here is my initial whack at it: Source Node: virsh # list IdName State 1 sl63 running virsh # migrate --p2p sl63 qemu+ssh://192.168.1.2/system error: operation failed: Failed to connect to remote libvirt URI qemu+ssh://3.57.111.32/system virsh # migrate --live sl63 qemu+ssh://192.168.1.2/system The authenticity of host '3.57.111.32 (192.168.1.2)' can't be established. RSA key fingerprint is e5:1d:b3:e5:38:5f:e1:8b:73:26:9e:15:c8:0a:2d:ac. Are you sure you want to continue connecting (yes/no)? yes root@192.168.1.2's password: Please enter your authentication name: vdsm@ovirt Please enter your password: virsh # Dest Node After migrate --live: virsh # list IdName State 2 sl63 running virsh # On Thu, Jan 31, 2013 at 10:38 AM, Dead Horse deadhorseconsult...@gmail.com wrote: Shu, I build oVirt Engine and vdsm from source myself. The commits I indicated are what I built from.I run the engine under FC17 and my nodes are running EL6.x respectively. Dan, I reverted VDSM on my two test nodes to an earlier build of VDSM (commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4) VDSM after the above commit is broken due to commit: fc3a44f71d2ef202cff18d7203b9e4165b546621 however when I built and tested from master yesterday I did apply a patch I tested for ybronhei which fixed that issue. I will build VDSM from master, today w/ the supervdsm patch and try the manual migration you indicated. - DHC On Thu, Jan 31, 2013 at 4:56 AM, Dan Kenigsberg dan...@redhat.comwrote: On Thu, Jan 31, 2013 at 11:08:58AM +0100, Martin Kletzander wrote: On 01/31/2013 10:25 AM, Dan Kenigsberg wrote: On Thu, Jan 31, 2013 at 09:43:44AM +0100, Martin Kletzander wrote: On 01/30/2013 08:40 PM, Dead Horse wrote: The nodes are EL6.3 based. Currently installed libvirt packages: libvirt-lock-sanlock-0.9.10-21.el6_3.8.x86_64 libvirt-cim-0.6.1-3.el6.x86_64 libvirt-0.9.10-21.el6_3.8.x86_64 libvirt-python-0.9.10-21.el6_3.8.x86_64 libvirt-client-0.9.10-21.el6_3.8.x86_64 and qemu packages: qemu-kvm-0.12.1.2-2.295.el6_3.10.x86_64 qemu-kvm-tools-0.12.1.2-2.295.el6_3.10.x86_64 qemu-img-0.12.1.2-2.295.el6_3.10.x86_64 Thus my presumption here given the above is that virDomainMigrateToURI2 has not yet been patched and/or back-ported into the EL6.x libvirt/qemu? virDomainMigrateToURI2 is supported since 0.9.2, but is there a possibility the code is requesting direct migration? That might explain the message, which is then incorrect; this was fixed in [1]. Martin [1] http://libvirt.org/git/?p=libvirt.git;a=commitdiff;h=3189dfb1636da22d426d2fc07cc9f60304b16c5c What is direct migration exactly, in the context of qemu-kvm? We are using p2p migration http://gerrit.ovirt.org/gitweb?p=vdsm.git;a=blob;f=vdsm/libvirtvm.py;h=fe140ecbfac665248e2ad5c4bfaebaf54ab884cc;hb=18c24f7c7c27ac732c4a760caa9524e7319cd47e#l501 OK, so that's not the issue, sorry for the confusion. I was thinking it would somehow get there. Direct migration doesn't exist in QEMU at all, so it seemed weird, but I can't seem to find any other reason for this failure; will keep searching, though. In this case, Dead Horse, would you try to migrate a VM (that you do not care much about) using virsh -c qemu+tls://hostname/system migrate --p2p dsthost? I'd like to see that the problem reproduces this way, too. More of libvirtd.log may help. You may want to disable iptables for a moment, just to eliminate a common cause of failure. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
18:03:04.500+: 2825: debug : virDomainInterfaceStats:7223 : dom=0x7f4f8c2455d0, (VM: name=sl63, uuid=887d764a-f835-4112-9eda-836a772ea5eb), path=vnet0, stats=0x7f4f9b6cdb00, size=64 2013-01-31 18:03:04.500+: 2825: debug : virDomainFree:2313 : dom=0x7f4f8c2455d0, (VM: name=sl63, uuid=887d764a-f835-4112-9eda-836a772ea5eb) 2013-01-31 18:03:09.505+: 2829: debug : virDomainInterfaceStats:7223 : dom=0x7f4f800034a0, (VM: name=sl63, uuid=887d764a-f835-4112-9eda-836a772ea5eb), path=vnet0, stats=0x7f4f98ec9b00, size=64 2013-01-31 18:03:09.520+: 2829: debug : virDomainFree:2313 : dom=0x7f4f800034a0, (VM: name=sl63, uuid=887d764a-f835-4112-9eda-836a772ea5eb) On Thu, Jan 31, 2013 at 11:27 AM, Dead Horse deadhorseconsult...@gmail.comwrote: note ignore the IP diff in the ssh host auth -- copy/paste fail ;) - DHC On Thu, Jan 31, 2013 at 11:25 AM, Dead Horse deadhorseconsult...@gmail.com wrote: Doh, brain fart VDSM is not involved here for the purposed of the needed test. Here is my initial whack at it: Source Node: virsh # list IdName State 1 sl63 running virsh # migrate --p2p sl63 qemu+ssh://192.168.1.2/system error: operation failed: Failed to connect to remote libvirt URI qemu+ssh://192.168.1.2/system http://3.57.111.32/system virsh # migrate --live sl63 qemu+ssh://192.168.1.2/system The authenticity of host '192.168.1.2 (192.168.1.2)' can't be established. RSA key fingerprint is e5:1d:b3:e5:38:5f:e1:8b:73:26:9e:15:c8:0a:2d:ac. Are you sure you want to continue connecting (yes/no)? yes root@192.168.1.2's password: Please enter your authentication name: vdsm@ovirt Please enter your password: virsh # Dest Node After migrate --live: virsh # list IdName State 2 sl63 running virsh # On Thu, Jan 31, 2013 at 10:38 AM, Dead Horse deadhorseconsult...@gmail.com wrote: Shu, I build oVirt Engine and vdsm from source myself. The commits I indicated are what I built from.I run the engine under FC17 and my nodes are running EL6.x respectively. Dan, I reverted VDSM on my two test nodes to an earlier build of VDSM (commit: c343e1833f7b6e5428dd90f14f7807dca1baa0b4) VDSM after the above commit is broken due to commit: fc3a44f71d2ef202cff18d7203b9e4165b546621 however when I built and tested from master yesterday I did apply a patch I tested for ybronhei which fixed that issue. I will build VDSM from master, today w/ the supervdsm patch and try the manual migration you indicated. - DHC On Thu, Jan 31, 2013 at 4:56 AM, Dan Kenigsberg dan...@redhat.comwrote: On Thu, Jan 31, 2013 at 11:08:58AM +0100, Martin Kletzander wrote: On 01/31/2013 10:25 AM, Dan Kenigsberg wrote: On Thu, Jan 31, 2013 at 09:43:44AM +0100, Martin Kletzander wrote: On 01/30/2013 08:40 PM, Dead Horse wrote: The nodes are EL6.3 based. Currently installed libvirt packages: libvirt-lock-sanlock-0.9.10-21.el6_3.8.x86_64 libvirt-cim-0.6.1-3.el6.x86_64 libvirt-0.9.10-21.el6_3.8.x86_64 libvirt-python-0.9.10-21.el6_3.8.x86_64 libvirt-client-0.9.10-21.el6_3.8.x86_64 and qemu packages: qemu-kvm-0.12.1.2-2.295.el6_3.10.x86_64 qemu-kvm-tools-0.12.1.2-2.295.el6_3.10.x86_64 qemu-img-0.12.1.2-2.295.el6_3.10.x86_64 Thus my presumption here given the above is that virDomainMigrateToURI2 has not yet been patched and/or back-ported into the EL6.x libvirt/qemu? virDomainMigrateToURI2 is supported since 0.9.2, but is there a possibility the code is requesting direct migration? That might explain the message, which is then incorrect; this was fixed in [1]. Martin [1] http://libvirt.org/git/?p=libvirt.git;a=commitdiff;h=3189dfb1636da22d426d2fc07cc9f60304b16c5c What is direct migration exactly, in the context of qemu-kvm? We are using p2p migration http://gerrit.ovirt.org/gitweb?p=vdsm.git;a=blob;f=vdsm/libvirtvm.py;h=fe140ecbfac665248e2ad5c4bfaebaf54ab884cc;hb=18c24f7c7c27ac732c4a760caa9524e7319cd47e#l501 OK, so that's not the issue, sorry for the confusion. I was thinking it would somehow get there. Direct migration doesn't exist in QEMU at all, so it seemed weird, but I can't seem to find any other reason for this failure; will keep searching, though. In this case, Dead Horse, would you try to migrate a VM (that you do not care much about) using virsh -c qemu+tls://hostname/system migrate --p2p dsthost? I'd like to see that the problem reproduces this way, too. More of libvirtd.log may help. You may want to disable iptables for a moment, just to eliminate a common cause of failure. ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] VM migrations failing
Engine Build -- Commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b
VDSM Build: -- Commit: da89a27492cc7d5a84e4bb87652569ca8e0fb20e + patch
-- http://gerrit.ovirt.org/#/c/11492/
Engine Side:
2013-01-30 10:56:38,439 ERROR
[org.ovirt.engine.core.vdsbroker.VdsUpdateRunTimeInfo]
(QuartzScheduler_Worker-70) Rerun vm 887d764a-f835-4112-9eda-836a772ea5eb.
Called from vds lostisles
2013-01-30 10:56:38,506 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) START, MigrateStatusVDSCommand(HostName = lostisles,
HostId = e042b03b-dd4e-414c-be1a-b2c65ac000f5,
vmId=887d764a-f835-4112-9eda-836a772ea5eb), log id: 6556e75b
2013-01-30 10:56:38,510 ERROR
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Failed in MigrateStatusVDS method
2013-01-30 10:56:38,510 ERROR
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Error code migrateErr and error message
VDSGenericException: VDSErrorException: Failed to MigrateStatusVDS, error =
Fatal error during migration
2013-01-30 10:56:38,511 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Command
org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand return
value
StatusOnlyReturnForXmlRpc [mStatus=StatusForXmlRpc [mCode=12,
mMessage=Fatal error during migration]]
VDSM Side:
Thread-43670::ERROR::2013-01-30 10:56:37,052::vm::200::vm.Vm::(_recover)
vmId=`887d764a-f835-4112-9eda-836a772ea5eb`::this function is not supported
by the connection driver: virDomainMigrateToURI2
Thread-43670::ERROR::2013-01-30 10:56:37,513::vm::288::vm.Vm::(run)
vmId=`887d764a-f835-4112-9eda-836a772ea5eb`::Failed to migrate
Traceback (most recent call last):
File /usr/share/vdsm/vm.py, line 273, in run
self._startUnderlyingMigration()
File /usr/share/vdsm/libvirtvm.py, line 504, in
_startUnderlyingMigration
None, maxBandwidth)
File /usr/share/vdsm/libvirtvm.py, line 540, in f
ret = attr(*args, **kwargs)
File /usr/lib64/python2.6/site-packages/vdsm/libvirtconnection.py, line
111, in wrapper
ret = f(*args, **kwargs)
File /usr/lib64/python2.6/site-packages/libvirt.py, line 1103, in
migrateToURI2
if ret == -1: raise libvirtError ('virDomainMigrateToURI2() failed',
dom=self)
libvirtError: this function is not supported by the connection driver:
virDomainMigrateToURI2
GuestMonitor-sl63::DEBUG::2013-01-30
10:56:38,235::libvirtvm::307::vm.Vm::(_getDiskLatency)
vmId=`887d764a-f835-4112-9eda-836a772ea5eb`::Disk vda latency not available
- DHC
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Cannot create vm's from user portal
That was indeed it ;) On Wed, Jan 30, 2013 at 8:10 AM, Eli Mesika emes...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Tuesday, January 29, 2013 5:24:20 AM Subject: Re: [Users] Cannot create vm's from user portal Going to venture a guess that this -- http://gerrit.ovirt.org/#/c/11411/ is the fix for this ;) Yes, it is !! - DHC On Mon, Jan 28, 2013 at 3:40 PM, Dead Horse deadhorseconsult...@gmail.com wrote: Seeing this error in engine log when attempting to create a new VM from the userportal: 2013-01-28 15:35:47,158 ERROR [org.ovirt.engine.core.bll.GetClustersWithPermittedActionQuery] (ajp--127.0.0.1-8702-5) Query GetClustersWithPermittedActionQuery failed. Exception message is PreparedStatementCallback; bad SQL grammar [select * from fn_perms_get_vds_groups_with_permitted_action(?, ?)]; nested exception is org.postgresql.util.PSQLException: ERROR: missing FROM-clause entry for table vds_groups Where: PL/pgSQL function fn_perms_get_vds_groups_with_permitted_action line 3 at RETURN QUERY 2013-01-28 15:36:16,726 ERROR [org.ovirt.engine.core.bll.GetClustersWithPermittedActionQuery] (ajp--127.0.0.1-8702-7) Query GetClustersWithPermittedActionQuery failed. Exception message is PreparedStatementCallback; bad SQL grammar [select * from fn_perms_get_vds_groups_with_permitted_action(?, ?)]; nested exception is org.postgresql.util.PSQLException: ERROR: missing FROM-clause entry for table vds_groups Where: PL/pgSQL function fn_perms_get_vds_groups_with_permitted_action line 3 at RETURN QUERY Current running engine build -- commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] VM migrations failing
The nodes are EL6.3 based.
Currently installed libvirt packages:
libvirt-lock-sanlock-0.9.10-21.el6_3.8.x86_64
libvirt-cim-0.6.1-3.el6.x86_64
libvirt-0.9.10-21.el6_3.8.x86_64
libvirt-python-0.9.10-21.el6_3.8.x86_64
libvirt-client-0.9.10-21.el6_3.8.x86_64
and qemu packages:
qemu-kvm-0.12.1.2-2.295.el6_3.10.x86_64
qemu-kvm-tools-0.12.1.2-2.295.el6_3.10.x86_64
qemu-img-0.12.1.2-2.295.el6_3.10.x86_64
Thus my presumption here given the above is that virDomainMigrateToURI2 has
not yet been patched and/or back-ported into the EL6.x libvirt/qemu?
- DHC
On Wed, Jan 30, 2013 at 1:28 PM, Dan Kenigsberg dan...@redhat.com wrote:
On Wed, Jan 30, 2013 at 11:04:00AM -0600, Dead Horse wrote:
Engine Build -- Commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b
VDSM Build: -- Commit: da89a27492cc7d5a84e4bb87652569ca8e0fb20e + patch
-- http://gerrit.ovirt.org/#/c/11492/
Engine Side:
2013-01-30 10:56:38,439 ERROR
[org.ovirt.engine.core.vdsbroker.VdsUpdateRunTimeInfo]
(QuartzScheduler_Worker-70) Rerun vm
887d764a-f835-4112-9eda-836a772ea5eb.
Called from vds lostisles
2013-01-30 10:56:38,506 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) START, MigrateStatusVDSCommand(HostName = lostisles,
HostId = e042b03b-dd4e-414c-be1a-b2c65ac000f5,
vmId=887d764a-f835-4112-9eda-836a772ea5eb), log id: 6556e75b
2013-01-30 10:56:38,510 ERROR
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Failed in MigrateStatusVDS method
2013-01-30 10:56:38,510 ERROR
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Error code migrateErr and error message
VDSGenericException: VDSErrorException: Failed to MigrateStatusVDS,
error =
Fatal error during migration
2013-01-30 10:56:38,511 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand]
(pool-3-thread-49) Command
org.ovirt.engine.core.vdsbroker.vdsbroker.MigrateStatusVDSCommand return
value
StatusOnlyReturnForXmlRpc [mStatus=StatusForXmlRpc [mCode=12,
mMessage=Fatal error during migration]]
VDSM Side:
Thread-43670::ERROR::2013-01-30 10:56:37,052::vm::200::vm.Vm::(_recover)
vmId=`887d764a-f835-4112-9eda-836a772ea5eb`::this function is not
supported
by the connection driver: virDomainMigrateToURI2
Thread-43670::ERROR::2013-01-30 10:56:37,513::vm::288::vm.Vm::(run)
vmId=`887d764a-f835-4112-9eda-836a772ea5eb`::Failed to migrate
Traceback (most recent call last):
File /usr/share/vdsm/vm.py, line 273, in run
self._startUnderlyingMigration()
File /usr/share/vdsm/libvirtvm.py, line 504, in
_startUnderlyingMigration
None, maxBandwidth)
File /usr/share/vdsm/libvirtvm.py, line 540, in f
ret = attr(*args, **kwargs)
File /usr/lib64/python2.6/site-packages/vdsm/libvirtconnection.py,
line
111, in wrapper
ret = f(*args, **kwargs)
File /usr/lib64/python2.6/site-packages/libvirt.py, line 1103, in
migrateToURI2
if ret == -1: raise libvirtError ('virDomainMigrateToURI2() failed',
dom=self)
libvirtError: this function is not supported by the connection driver:
virDomainMigrateToURI2
Could it be that you are using an ancient libvirt with no
virDomainMigrateToURI2? What are your libvirt and qemu-kvm versions (on
both machines)?
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
[Users] Engine upgrade failure broken in master
Commit: 72a51f5e21f38bf259a460948670eac92e97ca24 Breaks engine upgrades: 2013-01-29 12:40:52::ERROR::engine-upgrade::1177::root:: Traceback (most recent call last): File /usr/bin/engine-upgrade, line 1170, in module main(options) File /usr/bin/engine-upgrade, line 1071, in main if zombieTasksFound(): File /usr/bin/engine-upgrade, line 766, in zombieTasksFound msg=Can't get zombie async tasks, File /usr/share/ovirt-engine/scripts/common_utils.py, line 459, in execCmd env=env, File /usr/lib64/python2.7/subprocess.py, line 679, in __init__ errread, errwrite) File /usr/lib64/python2.7/subprocess.py, line 1249, in _execute_child raise child_exception OSError: [Errno 8] Exec format error Upgrade was attempted on engine built from commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 Building from commit: 82bdc46dfdb46b000f67f0cd4e51fc39665bf13b and upgrading works as expected. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] host power managment failures
Current running engine build -- commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 Host power management is having some issues: 2013-01-28 13:15:49,320 ERROR [org.ovirt.engine.core.bll.FenceExecutor] (ajp--127.0.0.1-8702-11) Illegal value in PM Proxy Preferences string , skipped. 2013-01-28 13:15:49,321 ERROR [org.ovirt.engine.core.bll.FenceExecutor] (ajp--127.0.0.1-8702-11) Failed to run Power Management command on Host , no running proxy Host was found. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] engine Failed to decrypt Data error
Password length is 11 characters and consists of Upper, Lower case and one special character. On Mon, Jan 28, 2013 at 1:20 PM, Alon Bar-Lev alo...@redhat.com wrote: We tried to reproduce this. What password do you use? is there one with some great length? If not, Eli, we should send a debug patch for this. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Monday, January 28, 2013 9:16:20 PM Subject: [Users] engine Failed to decrypt Data error I see this repeating error in the engine logs quite a bit, any ideas on what causes it? 2013-01-28 13:13:40,483 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-23) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:13:52,747 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-81) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:13:52,747 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-84) Failed to decrypt Blocktype mismatch: 0 2013-01-28 13:13:52,761 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-85) Failed to decrypt Data must start with zero 2013-01-28 13:14:00,964 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-23) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:00,964 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-20) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:02,983 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-29) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:02,983 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-34) Failed to decrypt Data must not be longer than 256 bytes - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] engine Failed to decrypt Data error
Current running engine build -- commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 ran engine upgrade against the built rpms from that commit. Thus I applied it as an upgrade against prior running build -- commit: 1eb895355239bbcb7a7ceda172405f0b68f18f35 On Mon, Jan 28, 2013 at 1:28 PM, Alon Bar-Lev alo...@redhat.com wrote: How do you installed the engine? you built? Which exact version? - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Alon Bar-Lev alo...@redhat.com Cc: users users@ovirt.org, Eli Mesika emes...@redhat.com Sent: Monday, January 28, 2013 9:26:44 PM Subject: Re: [Users] engine Failed to decrypt Data error Password length is 11 characters and consists of Upper, Lower case and one special character. On Mon, Jan 28, 2013 at 1:20 PM, Alon Bar-Lev alo...@redhat.com wrote: We tried to reproduce this. What password do you use? is there one with some great length? If not, Eli, we should send a debug patch for this. - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Monday, January 28, 2013 9:16:20 PM Subject: [Users] engine Failed to decrypt Data error I see this repeating error in the engine logs quite a bit, any ideas on what causes it? 2013-01-28 13:13:40,483 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-23) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:13:52,747 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-81) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:13:52,747 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-84) Failed to decrypt Blocktype mismatch: 0 2013-01-28 13:13:52,761 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-85) Failed to decrypt Data must start with zero 2013-01-28 13:14:00,964 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-23) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:00,964 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-20) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:02,983 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-29) Failed to decrypt Data must not be longer than 256 bytes 2013-01-28 13:14:02,983 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (QuartzScheduler_Worker-34) Failed to decrypt Data must not be longer than 256 bytes - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] host power managment failures
Removing and re-adding the host then configuring power management resolves this. - DHC On Mon, Jan 28, 2013 at 1:25 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Current running engine build -- commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 Host power management is having some issues: 2013-01-28 13:15:49,320 ERROR [org.ovirt.engine.core.bll.FenceExecutor] (ajp--127.0.0.1-8702-11) Illegal value in PM Proxy Preferences string , skipped. 2013-01-28 13:15:49,321 ERROR [org.ovirt.engine.core.bll.FenceExecutor] (ajp--127.0.0.1-8702-11) Failed to run Power Management command on Host , no running proxy Host was found. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] engine Failed to decrypt Data error
was in the middle of a fresh engine setup which did not exhibit the
symptom. However after running: engine-config -s
AdminPassword=interactive and restarting the engine service on the clean
setup the error message now shows up.
- DHC
On Mon, Jan 28, 2013 at 1:55 PM, Alon Bar-Lev alo...@redhat.com wrote:
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: Alon Bar-Lev alo...@redhat.com
Cc: users users@ovirt.org, Eli Mesika emes...@redhat.com
Sent: Monday, January 28, 2013 9:46:53 PM
Subject: Re: [Users] engine Failed to decrypt Data error
Current running engine build -- commit:
61c11aecc40e755d08b6c34c6fe1c0a07fa94de8
ran engine upgrade against the built rpms from that commit.
Thus I applied it as an upgrade against prior running build --
commit:
1eb895355239bbcb7a7ceda172405f0b68f18f35
[Please use plain text mails in lists.]
Can you please patch EncryptionUtils.decrypt() with the following, so I
can see what source is? source is encrypted blob, should not be a problem
to send it.
if (!StringHelper.isNullOrEmpty(source.trim())) {
KeyStore store = EncryptionUtils.getKeyStore(keyFile,
passwd, certType);
Key key = store.getKey(alias, passwd.toCharArray());
+log.info(DEBUG001 + source);
result = decrypt(source, key);
}
On Mon, Jan 28, 2013 at 1:28 PM, Alon Bar-Lev alo...@redhat.com
wrote:
How do you installed the engine? you built?
Which exact version?
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: Alon Bar-Lev alo...@redhat.com
Cc: users users@ovirt.org , Eli Mesika emes...@redhat.com
Sent: Monday, January 28, 2013 9:26:44 PM
Subject: Re: [Users] engine Failed to decrypt Data error
Password length is 11 characters and consists of Upper, Lower case
and one special character.
On Mon, Jan 28, 2013 at 1:20 PM, Alon Bar-Lev alo...@redhat.com
wrote:
We tried to reproduce this.
What password do you use? is there one with some great length?
If not, Eli, we should send a debug patch for this.
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: users@ovirt.org users@ovirt.org
Sent: Monday, January 28, 2013 9:16:20 PM
Subject: [Users] engine Failed to decrypt Data error
I see this repeating error in the engine logs quite a bit, any
ideas
on what causes it?
2013-01-28 13:13:40,483 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-23) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:13:52,747 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-81) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:13:52,747 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-84) Failed to decrypt Blocktype mismatch:
0
2013-01-28 13:13:52,761 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-85) Failed to decrypt Data must start
with
zero
2013-01-28 13:14:00,964 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-23) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:14:00,964 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-20) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:14:02,983 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-29) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:14:02,983 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-34) Failed to decrypt Data must not be
longer than 256 bytes
- DHC
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] engine Failed to decrypt Data error
psql -U engine -d engine -c select * from vdc_options where option_name
in ('LocalAdminPassword', 'AdminPassword');
option_id |option_name
|
option_value
| version
---++---
---+-
127 | LocalAdminPassword |
KiG8670o1qXVX6omYsiCdaaXtQc/mGmr0qgLHqc8yykoRz
OwbfZzU9AxBYwYrJEwyqdq8c2ZwfGVvQ1YVIfGRspKLKogl59gBnwcQuk3al1K4Vtmr2hgWDtm5FBYd5
Nac4WIly4efjMCRjwrpPVkpAX55N8tGJ9LNzX8eRszQ4iVs8zivl0eu9SVhrB8tbHkA/+U5/vss26za8
X+AV67dtDzoD7ZS0eOT1Vx9vrOGHvDYU8tANEb29Et79CJ0whLOOEeuwTpkK1yZdF3PaWRbnTwXZUsB1
hMs9NLdo2ZxZOVSIK1E2mPh1WLybgIX1YB0Ra3BZvjAR9wPZz+jdfZng== | general
7 | AdminPassword |
AakmoHu69RmCWkSoVXLOv0cwzwGscXaM+HJAONRtSdECEA
VL+bjc1Lis6PHR1vBwdmhITxAvo2998pTJNusvtuTCODra40MTC+9p9+Oev4jWIbkncHH8gRdIKyvHuz
O6fNda50VXeWYhGNFIMavw15PlslutUWEpyNAasjEWyZ7cNyjKK2eFKNDZ3F5PCv9RcQXfXkKSveWm6M
40zUVOx1ZjCnptNUpB4VYf5vW8LOpSL5NJpfJQmu36QbBRDDo3+3XPb4ELXA4t1rbPYw9Z7hRbk5Mbtq
qvOA7q4+G4nPtxHB7d6dYT2QJ58wgXUSIIoz/odvz5yVYeazIFS3Faww== | general
(2 rows)
On Mon, Jan 28, 2013 at 2:38 PM, Alon Bar-Lev alo...@redhat.com wrote:
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: Alon Bar-Lev alo...@redhat.com
Cc: users users@ovirt.org, Eli Mesika emes...@redhat.com
Sent: Monday, January 28, 2013 10:35:34 PM
Subject: Re: [Users] engine Failed to decrypt Data error
was in the middle of a fresh engine setup which did not exhibit the
symptom. However after running: engine-config -s
AdminPassword=interactive and restarting the engine service on the
clean setup the error message now shows up.
- DHC
OK, at least it is related to the admin password.
Please send me the output of:
psql -U engine -d engine -c select * from vdc_options where option_name
in ('LocalAdminPassword', 'AdminPassword');
Thanks!
On Mon, Jan 28, 2013 at 1:55 PM, Alon Bar-Lev alo...@redhat.com
wrote:
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: Alon Bar-Lev alo...@redhat.com
Cc: users users@ovirt.org , Eli Mesika emes...@redhat.com
Sent: Monday, January 28, 2013 9:46:53 PM
Subject: Re: [Users] engine Failed to decrypt Data error
Current running engine build -- commit:
61c11aecc40e755d08b6c34c6fe1c0a07fa94de8
ran engine upgrade against the built rpms from that commit.
Thus I applied it as an upgrade against prior running build --
commit:
1eb895355239bbcb7a7ceda172405f0b68f18f35
[Please use plain text mails in lists.]
Can you please patch EncryptionUtils.decrypt() with the following, so
I can see what source is? source is encrypted blob, should not be a
problem to send it.
if (!StringHelper.isNullOrEmpty(source.trim())) {
KeyStore store = EncryptionUtils.getKeyStore(keyFile, passwd,
certType);
Key key = store.getKey(alias, passwd.toCharArray());
+ log.info (DEBUG001 + source);
result = decrypt(source, key);
}
On Mon, Jan 28, 2013 at 1:28 PM, Alon Bar-Lev alo...@redhat.com
wrote:
How do you installed the engine? you built?
Which exact version?
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: Alon Bar-Lev alo...@redhat.com
Cc: users users@ovirt.org , Eli Mesika
emes...@redhat.com
Sent: Monday, January 28, 2013 9:26:44 PM
Subject: Re: [Users] engine Failed to decrypt Data error
Password length is 11 characters and consists of Upper, Lower
case
and one special character.
On Mon, Jan 28, 2013 at 1:20 PM, Alon Bar-Lev alo...@redhat.com
wrote:
We tried to reproduce this.
What password do you use? is there one with some great length?
If not, Eli, we should send a debug patch for this.
- Original Message -
From: Dead Horse deadhorseconsult...@gmail.com
To: users@ovirt.org users@ovirt.org
Sent: Monday, January 28, 2013 9:16:20 PM
Subject: [Users] engine Failed to decrypt Data error
I see this repeating error in the engine logs quite a bit, any
ideas
on what causes it?
2013-01-28 13:13:40,483 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-23) Failed to decrypt Data must not be
longer than 256 bytes
2013-01-28 13:13:52,747 ERROR
[org.ovirt.engine.core.engineencryptutils.EncryptionUtils]
(QuartzScheduler_Worker-81) Failed to decrypt Data must
Re: [Users] Cannot create vm's from user portal
Going to venture a guess that this -- http://gerrit.ovirt.org/#/c/11411/is the fix for this ;) - DHC On Mon, Jan 28, 2013 at 3:40 PM, Dead Horse deadhorseconsult...@gmail.comwrote: Seeing this error in engine log when attempting to create a new VM from the userportal: 2013-01-28 15:35:47,158 ERROR [org.ovirt.engine.core.bll.GetClustersWithPermittedActionQuery] (ajp--127.0.0.1-8702-5) Query GetClustersWithPermittedActionQuery failed. Exception message is PreparedStatementCallback; bad SQL grammar [select * from fn_perms_get_vds_groups_with_permitted_action(?, ?)]; nested exception is org.postgresql.util.PSQLException: ERROR: missing FROM-clause entry for table vds_groups Where: PL/pgSQL function fn_perms_get_vds_groups_with_permitted_action line 3 at RETURN QUERY 2013-01-28 15:36:16,726 ERROR [org.ovirt.engine.core.bll.GetClustersWithPermittedActionQuery] (ajp--127.0.0.1-8702-7) Query GetClustersWithPermittedActionQuery failed. Exception message is PreparedStatementCallback; bad SQL grammar [select * from fn_perms_get_vds_groups_with_permitted_action(?, ?)]; nested exception is org.postgresql.util.PSQLException: ERROR: missing FROM-clause entry for table vds_groups Where: PL/pgSQL function fn_perms_get_vds_groups_with_permitted_action line 3 at RETURN QUERY Current running engine build -- commit: 61c11aecc40e755d08b6c34c6fe1c0a07fa94de8 - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] latest vdsm cannot read ib device speeds causing storage attach fail
This test harness setup here consists of two servers tied to NFS storage
via IB (NFS mounts are via IPoIB, NFS over RDMA is disabled) . All storage
domains are NFS. The issue does occur with both servers on when attempting
to bring them out of maintenance mode with the end result being
non-operational due to storage attach fail.
The current issue is now that with a working older commit the master
storage domain is stuck in state locked and I see the secondary issue
wherein VDSM cannot seem to find or contact the master storage domain even
though it is there. I am can mount the master storage domain manually and
and all content appears to be accounted for accordingly on either host.
Here is the current contents of the master storage domain metadata:
CLASS=Data
DESCRIPTION=orgrimmar
IOOPTIMEOUTSEC=1
LEASERETRIES=3
LEASETIMESEC=5
LOCKPOLICY=
LOCKRENEWALINTERVALSEC=5
MASTER_VERSION=417
POOL_DESCRIPTION=Azeroth
POOL_DOMAINS=0549ee91-4498-4130-8c23-4c173b5c0959:Active,d8b55105-c90a-465d-9803-8130da9a671e:Active,67534cca-1327-462a-b455-a04464084b31:Active,c331a800-839d-4d23-9059-870a7471240a:Active,f8984825-ff8d-43d9-91db-0d0959f8bae9:Active,c434056e-96be-4702-8beb-82a408a5c8cb:Active,f7da73c7-b5fe-48b6-93a0-0c773018c94f:Active,82e3b34a-6f89-4299-8cd8-2cc8f973a3b4:Active,e615c975-6b00-469f-8fb6-ff58ae3fdb2c:Active,5bc86532-55f7-4a91-a52c-fad261f322d5:Active,1130b87a-3b34-45d6-8016-d435825c68ef:Active
POOL_SPM_ID=1
POOL_SPM_LVER=6
POOL_UUID=f90a0d1c-06ca-11e2-a05b-00151712f280
REMOTE_PATH=192.168.0.1:/ovirt/orgrimmar
ROLE=Master
SDUUID=67534cca-1327-462a-b455-a04464084b31
TYPE=NFS
VERSION=3
_SHA_CKSUM=1442bb078fd8c9468d241ff141e9bf53839f0721
So now with the older working commit I now get this the
StoragePoolMasterNotFound: Cannot find master domain error (prior details
above when I worked backwards to that commit)
This is odd as the nodes can definitely reach the master storage domain:
showmount from one of the el6.3 nodes:
[root@kezan ~]# showmount -e 192.168.0.1
Export list for 192.168.0.1:
/ovirt/orgrimmar192.168.0.0/16
mount/ls from one of the nodes:
[root@kezan ~]# mount 192.168.0.1:/ovirt/orgrimmar /mnt
[root@kezan ~]# ls -al /mnt/67534cca-1327-462a-b455-a04464084b31/dom_md/
total 1100
drwxr-xr-x 2 vdsm kvm4096 Jan 24 11:44 .
drwxr-xr-x 5 vdsm kvm4096 Oct 19 16:16 ..
-rw-rw 1 vdsm kvm 1048576 Jan 19 22:09 ids
-rw-rw 1 vdsm kvm 0 Sep 25 00:46 inbox
-rw-rw 1 vdsm kvm 2097152 Jan 10 13:33 leases
-rw-r--r-- 1 vdsm kvm 903 Jan 10 13:39 metadata
-rw-rw 1 vdsm kvm 0 Sep 25 00:46 outbox
- DHC
On Thu, Jan 24, 2013 at 7:51 AM, ybronhei ybron...@redhat.com wrote:
On 01/24/2013 12:44 AM, Dead Horse wrote:
I narrowed down on the commit where the originally reported issue crept
in:
commitfc3a44f71d2ef202cff18d72**03b9e4165b546621building and testing with
this commit or subsequent commits yields the original issue.
Interesting.. it might be related to this commit and we're trying to
reproduce it.
Did you try to remove that code and run again? does it work without the
additional of zombieReaper?
does the connectivity to the storage work well? when you run 'ls' on the
mounted folder you get see the files without a long delay ? it might
related to too long timeout when validating access to this mount..
we work on that.. any additional info can help
Thanks.
- DHC
On Wed, Jan 23, 2013 at 3:56 PM, Dead Horse
deadhorseconsult...@gmail.com**wrote:
Indeed reverting back to an older vdsm clears up the above issue. However
now I the issue is see is:
Thread-18::ERROR::2013-01-23
15:50:42,885::task::833::**TaskManager.Task::(_setError)
Task=`08709e68-bcbc-40d8-843a-**d69d4df40ac6`::Unexpected error
Traceback (most recent call last):
File /usr/share/vdsm/storage/task.**py, line 840, in _run
return fn(*args, **kargs)
File /usr/share/vdsm/logUtils.py, line 42, in wrapper
res = f(*args, **kwargs)
File /usr/share/vdsm/storage/hsm.**py, line 923, in
connectStoragePool
masterVersion, options)
File /usr/share/vdsm/storage/hsm.**py, line 970, in
_connectStoragePool
res = pool.connect(hostID, scsiKey, msdUUID, masterVersion)
File /usr/share/vdsm/storage/sp.**py, line 643, in connect
self.__rebuild(msdUUID=**msdUUID, masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.**py, line 1167, in __rebuild
self.masterDomain = self.getMasterDomain(msdUUID=**msdUUID,
masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.**py, line 1506, in getMasterDomain
raise se.StoragePoolMasterNotFound(**self.spUUID, msdUUID)
StoragePoolMasterNotFound: Cannot find master domain:
'spUUID=f90a0d1c-06ca-11e2-**a05b-00151712f280,
msdUUID=67534cca-1327-462a-**b455-a04464084b31'
Thread-18::DEBUG::2013-01-23
15:50:42,887::task::852::**TaskManager.Task::(_run)
Task=`08709e68-bcbc-40d8-843a-**d69d4df40ac6`::Task._run:
08709e68-bcbc-40d8-843a-**d69d4df40ac6
('f90a0d1c-06ca-11e2-a05b
Re: [Users] latest vdsm cannot read ib device speeds causing storage attach fail
Tried some manual edits to SD states in the dbase. The net result was I was able to get a node active. However as reconstructing the master storage domain kicked in it was unable to do so. It was also not able to recognize the other SD with similar failure modes to the unrecognized master above. Guessing the newer VDSM version borked things pretty good. So being as this is a test harness and the SD data is not worth saving I just smoked the all SD. I ran engine-cleanup and started from fresh and all is well now. - DHC On Thu, Jan 24, 2013 at 11:53 AM, Dead Horse deadhorseconsult...@gmail.comwrote: This test harness setup here consists of two servers tied to NFS storage via IB (NFS mounts are via IPoIB, NFS over RDMA is disabled) . All storage domains are NFS. The issue does occur with both servers on when attempting to bring them out of maintenance mode with the end result being non-operational due to storage attach fail. The current issue is now that with a working older commit the master storage domain is stuck in state locked and I see the secondary issue wherein VDSM cannot seem to find or contact the master storage domain even though it is there. I am can mount the master storage domain manually and and all content appears to be accounted for accordingly on either host. Here is the current contents of the master storage domain metadata: CLASS=Data DESCRIPTION=orgrimmar IOOPTIMEOUTSEC=1 LEASERETRIES=3 LEASETIMESEC=5 LOCKPOLICY= LOCKRENEWALINTERVALSEC=5 MASTER_VERSION=417 POOL_DESCRIPTION=Azeroth POOL_DOMAINS=0549ee91-4498-4130-8c23-4c173b5c0959:Active,d8b55105-c90a-465d-9803-8130da9a671e:Active,67534cca-1327-462a-b455-a04464084b31:Active,c331a800-839d-4d23-9059-870a7471240a:Active,f8984825-ff8d-43d9-91db-0d0959f8bae9:Active,c434056e-96be-4702-8beb-82a408a5c8cb:Active,f7da73c7-b5fe-48b6-93a0-0c773018c94f:Active,82e3b34a-6f89-4299-8cd8-2cc8f973a3b4:Active,e615c975-6b00-469f-8fb6-ff58ae3fdb2c:Active,5bc86532-55f7-4a91-a52c-fad261f322d5:Active,1130b87a-3b34-45d6-8016-d435825c68ef:Active POOL_SPM_ID=1 POOL_SPM_LVER=6 POOL_UUID=f90a0d1c-06ca-11e2-a05b-00151712f280 REMOTE_PATH=192.168.0.1:/ovirt/orgrimmar ROLE=Master SDUUID=67534cca-1327-462a-b455-a04464084b31 TYPE=NFS VERSION=3 _SHA_CKSUM=1442bb078fd8c9468d241ff141e9bf53839f0721 So now with the older working commit I now get this the StoragePoolMasterNotFound: Cannot find master domain error (prior details above when I worked backwards to that commit) This is odd as the nodes can definitely reach the master storage domain: showmount from one of the el6.3 nodes: [root@kezan ~]# showmount -e 192.168.0.1 Export list for 192.168.0.1: /ovirt/orgrimmar192.168.0.0/16 mount/ls from one of the nodes: [root@kezan ~]# mount 192.168.0.1:/ovirt/orgrimmar /mnt [root@kezan ~]# ls -al /mnt/67534cca-1327-462a-b455-a04464084b31/dom_md/ total 1100 drwxr-xr-x 2 vdsm kvm4096 Jan 24 11:44 . drwxr-xr-x 5 vdsm kvm4096 Oct 19 16:16 .. -rw-rw 1 vdsm kvm 1048576 Jan 19 22:09 ids -rw-rw 1 vdsm kvm 0 Sep 25 00:46 inbox -rw-rw 1 vdsm kvm 2097152 Jan 10 13:33 leases -rw-r--r-- 1 vdsm kvm 903 Jan 10 13:39 metadata -rw-rw 1 vdsm kvm 0 Sep 25 00:46 outbox - DHC On Thu, Jan 24, 2013 at 7:51 AM, ybronhei ybron...@redhat.com wrote: On 01/24/2013 12:44 AM, Dead Horse wrote: I narrowed down on the commit where the originally reported issue crept in: commitfc3a44f71d2ef202cff18d72**03b9e4165b546621building and testing with this commit or subsequent commits yields the original issue. Interesting.. it might be related to this commit and we're trying to reproduce it. Did you try to remove that code and run again? does it work without the additional of zombieReaper? does the connectivity to the storage work well? when you run 'ls' on the mounted folder you get see the files without a long delay ? it might related to too long timeout when validating access to this mount.. we work on that.. any additional info can help Thanks. - DHC On Wed, Jan 23, 2013 at 3:56 PM, Dead Horse deadhorseconsult...@gmail.com**wrote: Indeed reverting back to an older vdsm clears up the above issue. However now I the issue is see is: Thread-18::ERROR::2013-01-23 15:50:42,885::task::833::**TaskManager.Task::(_setError) Task=`08709e68-bcbc-40d8-843a-**d69d4df40ac6`::Unexpected error Traceback (most recent call last): File /usr/share/vdsm/storage/task.**py, line 840, in _run return fn(*args, **kargs) File /usr/share/vdsm/logUtils.py, line 42, in wrapper res = f(*args, **kwargs) File /usr/share/vdsm/storage/hsm.**py, line 923, in connectStoragePool masterVersion, options) File /usr/share/vdsm/storage/hsm.**py, line 970, in _connectStoragePool res = pool.connect(hostID, scsiKey, msdUUID, masterVersion) File /usr/share/vdsm/storage/sp.**py, line 643, in connect self.__rebuild(msdUUID=**msdUUID, masterVersion=masterVersion) File
Re: [Users] Encrypted Admin password error
On Tue, Jan 22, 2013 at 4:12 PM, Alon Bar-Lev alo...@redhat.com wrote: engine-config -s AdminPassword=interactive That did the trick vs passing the passwd via stdin directly from a command line. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] latest vdsm cannot read ib device speeds causing storage attach fail
Indeed reverting back to an older vdsm clears up the above issue. However
now I the issue is see is:
Thread-18::ERROR::2013-01-23
15:50:42,885::task::833::TaskManager.Task::(_setError)
Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Unexpected error
Traceback (most recent call last):
File /usr/share/vdsm/storage/task.py, line 840, in _run
return fn(*args, **kargs)
File /usr/share/vdsm/logUtils.py, line 42, in wrapper
res = f(*args, **kwargs)
File /usr/share/vdsm/storage/hsm.py, line 923, in connectStoragePool
masterVersion, options)
File /usr/share/vdsm/storage/hsm.py, line 970, in _connectStoragePool
res = pool.connect(hostID, scsiKey, msdUUID, masterVersion)
File /usr/share/vdsm/storage/sp.py, line 643, in connect
self.__rebuild(msdUUID=msdUUID, masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.py, line 1167, in __rebuild
self.masterDomain = self.getMasterDomain(msdUUID=msdUUID,
masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.py, line 1506, in getMasterDomain
raise se.StoragePoolMasterNotFound(self.spUUID, msdUUID)
StoragePoolMasterNotFound: Cannot find master domain:
'spUUID=f90a0d1c-06ca-11e2-a05b-00151712f280,
msdUUID=67534cca-1327-462a-b455-a04464084b31'
Thread-18::DEBUG::2013-01-23
15:50:42,887::task::852::TaskManager.Task::(_run)
Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Task._run:
08709e68-bcbc-40d8-843a-d69d4df40ac6
('f90a0d1c-06ca-11e2-a05b-00151712f280', 2,
'f90a0d1c-06ca-11e2-a05b-00151712f280',
'67534cca-1327-462a-b455-a04464084b31', 433) {} failed - stopping task
This is with vdsm built from
commit25a2d8572ad32352227c98a86631300fbd6523c1
- DHC
On Wed, Jan 23, 2013 at 10:44 AM, Dead Horse
deadhorseconsult...@gmail.comwrote:
VDSM was built from:
commit 166138e37e75767b32227746bb671b1dab9cdd5e
Attached is the full vdsm log
I should also note that from engine perspective it sees the master storage
domain as locked and the others as unknown.
On Wed, Jan 23, 2013 at 2:49 AM, Dan Kenigsberg dan...@redhat.com wrote:
On Tue, Jan 22, 2013 at 04:02:24PM -0600, Dead Horse wrote:
Any ideas on this one? (from VDSM log):
Thread-25::DEBUG::2013-01-22
15:35:29,065::BindingXMLRPC::914::vds::(wrapper) client
[3.57.111.30]::call
getCapabilities with () {}
Thread-25::ERROR::2013-01-22 15:35:29,113::netinfo::159::root::(speed)
cannot read ib0 speed
Traceback (most recent call last):
File /usr/lib64/python2.6/site-packages/vdsm/netinfo.py, line 155,
in
speed
s = int(file('/sys/class/net/%s/speed' % dev).read())
IOError: [Errno 22] Invalid argument
Causes VDSM to fail to attach storage
I doubt that this is the cause of the failure, as vdsm has always
reported 0 for ib devices, and still is.
Does a former version works with your Engine?
Could you share more of your vdsm.log? I suppose the culprit lies in one
one of the storage-related commands, not in statistics retrieval.
Engine side sees:
ERROR [org.ovirt.engine.core.bll.storage.NFSStorageHelper]
(QuartzScheduler_Worker-96) [553ef26e] The connection with details
192.168.0.1:/ovirt/ds failed because of error code 100 and error
message
is: general exception
2013-01-22 15:35:30,160 INFO
[org.ovirt.engine.core.bll.SetNonOperationalVdsCommand]
(QuartzScheduler_Worker-96) [1ab78378] Running command:
SetNonOperationalVdsCommand internal: true. Entities affected : ID:
8970b3fe-1faf-11e2-bc1f-00151712f280 Type: VDS
2013-01-22 15:35:30,200 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] START,
SetVdsStatusVDSCommand(HostName = kezan, HostId =
8970b3fe-1faf-11e2-bc1f-00151712f280, status=NonOperational,
nonOperationalReason=STORAGE_DOMAIN_UNREACHABLE), log id: 4af5c4cd
2013-01-22 15:35:30,211 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] FINISH, SetVdsStatusVDSCommand,
log
id: 4af5c4cd
2013-01-22 15:35:30,242 ERROR
[org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
(QuartzScheduler_Worker-96) [1ab78378] Try to add duplicate audit log
values with the same name. Type: VDS_SET_NONOPERATIONAL_DOMAIN. Value:
storagepoolname
Engine = latest master
VDSM = latest master
Since latest master is an unstable reference by definition, I'm sure
that History would thank you if you post the exact version (git hash?)
of the code.
node = el6
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] latest vdsm cannot read ib device speeds causing storage attach fail
I narrowed down on the commit where the originally reported issue crept in:
commitfc3a44f71d2ef202cff18d7203b9e4165b546621building and testing with
this commit or subsequent commits yields the original issue.
- DHC
On Wed, Jan 23, 2013 at 3:56 PM, Dead Horse
deadhorseconsult...@gmail.comwrote:
Indeed reverting back to an older vdsm clears up the above issue. However
now I the issue is see is:
Thread-18::ERROR::2013-01-23
15:50:42,885::task::833::TaskManager.Task::(_setError)
Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Unexpected error
Traceback (most recent call last):
File /usr/share/vdsm/storage/task.py, line 840, in _run
return fn(*args, **kargs)
File /usr/share/vdsm/logUtils.py, line 42, in wrapper
res = f(*args, **kwargs)
File /usr/share/vdsm/storage/hsm.py, line 923, in connectStoragePool
masterVersion, options)
File /usr/share/vdsm/storage/hsm.py, line 970, in _connectStoragePool
res = pool.connect(hostID, scsiKey, msdUUID, masterVersion)
File /usr/share/vdsm/storage/sp.py, line 643, in connect
self.__rebuild(msdUUID=msdUUID, masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.py, line 1167, in __rebuild
self.masterDomain = self.getMasterDomain(msdUUID=msdUUID,
masterVersion=masterVersion)
File /usr/share/vdsm/storage/sp.py, line 1506, in getMasterDomain
raise se.StoragePoolMasterNotFound(self.spUUID, msdUUID)
StoragePoolMasterNotFound: Cannot find master domain:
'spUUID=f90a0d1c-06ca-11e2-a05b-00151712f280,
msdUUID=67534cca-1327-462a-b455-a04464084b31'
Thread-18::DEBUG::2013-01-23
15:50:42,887::task::852::TaskManager.Task::(_run)
Task=`08709e68-bcbc-40d8-843a-d69d4df40ac6`::Task._run:
08709e68-bcbc-40d8-843a-d69d4df40ac6
('f90a0d1c-06ca-11e2-a05b-00151712f280', 2,
'f90a0d1c-06ca-11e2-a05b-00151712f280',
'67534cca-1327-462a-b455-a04464084b31', 433) {} failed - stopping task
This is with vdsm built from
commit25a2d8572ad32352227c98a86631300fbd6523c1
- DHC
On Wed, Jan 23, 2013 at 10:44 AM, Dead Horse
deadhorseconsult...@gmail.com wrote:
VDSM was built from:
commit 166138e37e75767b32227746bb671b1dab9cdd5e
Attached is the full vdsm log
I should also note that from engine perspective it sees the master
storage domain as locked and the others as unknown.
On Wed, Jan 23, 2013 at 2:49 AM, Dan Kenigsberg dan...@redhat.comwrote:
On Tue, Jan 22, 2013 at 04:02:24PM -0600, Dead Horse wrote:
Any ideas on this one? (from VDSM log):
Thread-25::DEBUG::2013-01-22
15:35:29,065::BindingXMLRPC::914::vds::(wrapper) client
[3.57.111.30]::call
getCapabilities with () {}
Thread-25::ERROR::2013-01-22 15:35:29,113::netinfo::159::root::(speed)
cannot read ib0 speed
Traceback (most recent call last):
File /usr/lib64/python2.6/site-packages/vdsm/netinfo.py, line 155,
in
speed
s = int(file('/sys/class/net/%s/speed' % dev).read())
IOError: [Errno 22] Invalid argument
Causes VDSM to fail to attach storage
I doubt that this is the cause of the failure, as vdsm has always
reported 0 for ib devices, and still is.
Does a former version works with your Engine?
Could you share more of your vdsm.log? I suppose the culprit lies in one
one of the storage-related commands, not in statistics retrieval.
Engine side sees:
ERROR [org.ovirt.engine.core.bll.storage.NFSStorageHelper]
(QuartzScheduler_Worker-96) [553ef26e] The connection with details
192.168.0.1:/ovirt/ds failed because of error code 100 and error
message
is: general exception
2013-01-22 15:35:30,160 INFO
[org.ovirt.engine.core.bll.SetNonOperationalVdsCommand]
(QuartzScheduler_Worker-96) [1ab78378] Running command:
SetNonOperationalVdsCommand internal: true. Entities affected : ID:
8970b3fe-1faf-11e2-bc1f-00151712f280 Type: VDS
2013-01-22 15:35:30,200 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] START,
SetVdsStatusVDSCommand(HostName = kezan, HostId =
8970b3fe-1faf-11e2-bc1f-00151712f280, status=NonOperational,
nonOperationalReason=STORAGE_DOMAIN_UNREACHABLE), log id: 4af5c4cd
2013-01-22 15:35:30,211 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] FINISH, SetVdsStatusVDSCommand,
log
id: 4af5c4cd
2013-01-22 15:35:30,242 ERROR
[org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
(QuartzScheduler_Worker-96) [1ab78378] Try to add duplicate audit log
values with the same name. Type: VDS_SET_NONOPERATIONAL_DOMAIN. Value:
storagepoolname
Engine = latest master
VDSM = latest master
Since latest master is an unstable reference by definition, I'm sure
that History would thank you if you post the exact version (git hash?)
of the code.
node = el6
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
[Users] latest vdsm cannot read ib device speeds causing storage attach fail
Any ideas on this one? (from VDSM log):
Thread-25::DEBUG::2013-01-22
15:35:29,065::BindingXMLRPC::914::vds::(wrapper) client [3.57.111.30]::call
getCapabilities with () {}
Thread-25::ERROR::2013-01-22 15:35:29,113::netinfo::159::root::(speed)
cannot read ib0 speed
Traceback (most recent call last):
File /usr/lib64/python2.6/site-packages/vdsm/netinfo.py, line 155, in
speed
s = int(file('/sys/class/net/%s/speed' % dev).read())
IOError: [Errno 22] Invalid argument
Causes VDSM to fail to attach storage
Engine side sees:
ERROR [org.ovirt.engine.core.bll.storage.NFSStorageHelper]
(QuartzScheduler_Worker-96) [553ef26e] The connection with details
192.168.0.1:/ovirt/ds failed because of error code 100 and error message
is: general exception
2013-01-22 15:35:30,160 INFO
[org.ovirt.engine.core.bll.SetNonOperationalVdsCommand]
(QuartzScheduler_Worker-96) [1ab78378] Running command:
SetNonOperationalVdsCommand internal: true. Entities affected : ID:
8970b3fe-1faf-11e2-bc1f-00151712f280 Type: VDS
2013-01-22 15:35:30,200 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] START,
SetVdsStatusVDSCommand(HostName = kezan, HostId =
8970b3fe-1faf-11e2-bc1f-00151712f280, status=NonOperational,
nonOperationalReason=STORAGE_DOMAIN_UNREACHABLE), log id: 4af5c4cd
2013-01-22 15:35:30,211 INFO
[org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand]
(QuartzScheduler_Worker-96) [1ab78378] FINISH, SetVdsStatusVDSCommand, log
id: 4af5c4cd
2013-01-22 15:35:30,242 ERROR
[org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
(QuartzScheduler_Worker-96) [1ab78378] Try to add duplicate audit log
values with the same name. Type: VDS_SET_NONOPERATIONAL_DOMAIN. Value:
storagepoolname
Engine = latest master
VDSM = latest master
node = el6
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users
[Users] Encrypted Admin password error
This error appears very frequently in the engine.log. 2013-01-22 15:33:10,837 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (MSC service thread 1-5) Failed to decrypt Data must start with zero 2013-01-22 15:33:10,838 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (MSC service thread 1-5) Failed to decrypt value for property TruststorePass will be used encrypted value 2013-01-22 15:33:10,865 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (MSC service thread 1-5) Failed to decrypt Data must start with zero 2013-01-22 15:33:10,866 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (MSC service thread 1-5) Failed to decrypt value for property AdminPassword will be used encrypted value Starting appearing after manual edit to the AdminPassword value in the database. Tried using engine-config -s AdminPassword='somepassword' to change it which always resulted in: cannot set value 'somepassword' to key AdminPassword. Hence the manual edit Any ideas on how to make the engine happy in that regard again? - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Latest VDSM uses newer qemu than EL6 provides
I figured that was the case, thanks for the heads up. - DHC On Fri, Nov 9, 2012 at 3:35 AM, Vinzenz Feenstra vfeen...@redhat.comwrote: Hi, Well this is a RPM spec file change, qemu-kvm will eventually follow as well after it was tested. It just needs some more time to be released. Regards, On 11/09/2012 03:32 AM, Dead Horse wrote: I noticed a change 12 days ago in the VDSM master branch with what version of qemu that VDSM requires. (Commit: 1ce73032d9238a1bf4c55400392d5ce47dfa2d43) Change: # Update the qemu-kvm requires when block_stream will be included -Requires: qemu-kvm = 2:0.12.1.2-2.227 -Requires: qemu-img = 2:0.12.1.2-2.227 +Requires: qemu-kvm = 2:0.12.1.2-2.295.el6_3.4 +Requires: qemu-img = 2:0.12.1.2-2.295.el6_3.4 However the version of qemu required has not yet materialized in any of the released EL sources/updates. - DHC ___ Users mailing listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users -- Regards, Vinzenz Feenstra | Senior Software Engineer RedHat Engineering Virtualization R D Phone: +420 532 294 625 IRC: vfeenstr or evilissimo Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Latest VDSM uses newer qemu than EL6 provides
I noticed a change 12 days ago in the VDSM master branch with what version of qemu that VDSM requires. (Commit: 1ce73032d9238a1bf4c55400392d5ce47dfa2d43) Change: # Update the qemu-kvm requires when block_stream will be included -Requires: qemu-kvm = 2:0.12.1.2-2.227 -Requires: qemu-img = 2:0.12.1.2-2.227 +Requires: qemu-kvm = 2:0.12.1.2-2.295.el6_3.4 +Requires: qemu-img = 2:0.12.1.2-2.295.el6_3.4 However the version of qemu required has not yet materialized in any of the released EL sources/updates. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] export domain issues with latest nightly/GIT Master
On issue 1 attached are log files narrowed down what was logged during the attempted import when the disks are not shown. On issue 2 I found this bug ( https://bugzilla.redhat.com/show_bug.cgi?id=801112) which seems to have a similar footprint. This was working a few weeks back so I would classify this as a regression. On Wed, Nov 7, 2012 at 2:59 AM, Eli Mesika emes...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: users@ovirt.org users@ovirt.org Sent: Wednesday, November 7, 2012 3:12:07 AM Subject: [Users] export domain issues with latest nightly/GIT Master I have noted some export domain issues with builds from the latest GIT Master. 1) When importing a VM, the disks subtab of the VM to import will not show the disks. Instead it shows the blinking progress squares. This is persistent and will it will not show information on the disks to be imported. This stopped working about 2 weeks back. Expected is to show the disks to import as well as set provisioning type and destination storage domain. Hi, can you plesae attach engin/vdsm logs as this is occuring generally when an exception is thrown in the middle of the import operation. 2) The import process does does not respect thin provisioning. Importing a VM without checking the collapse snapshots box (there are actually no snapshots to collapse...) results in the imported disks being imported thickly provisioned. If the collapse snapshots box is checked thin provisioning is respected and the disks import thin provisioned. This may be intertwined with the prior issue. Is that as regression or a new bug ? - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users 2012-11-07 21:22:45,978 INFO [org.ovirt.engine.core.bll.LoginAdminUserCommand] (ajp--127.0.0.1-8702-11) Checking if user admin@internal is an admin, result true 2012-11-07 21:22:45,980 INFO [org.ovirt.engine.core.bll.LoginAdminUserCommand] (ajp--127.0.0.1-8702-11) Running command: LoginAdminUserCommand internal: false. 2012-11-07 21:22:46,139 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt Data must start with zero 2012-11-07 21:22:46,140 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt value for property LocalAdminPassword will be used encrypted value 2012-11-07 21:22:46,162 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt Data must start with zero 2012-11-07 21:22:46,162 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt value for property LocalAdminPassword will be used encrypted value 2012-11-07 21:22:46,183 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt Data must start with zero 2012-11-07 21:22:46,183 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt value for property LocalAdminPassword will be used encrypted value 2012-11-07 21:22:46,204 ERROR [org.ovirt.engine.core.engineencryptutils.EncryptionUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt Data must start with zero 2012-11-07 21:22:46,205 ERROR [org.ovirt.engine.core.dal.dbbroker.generic.DBConfigUtils] (ajp--127.0.0.1-8702-1) Failed to decrypt value for property LocalAdminPassword will be used encrypted value 2012-11-07 21:22:51,085 ERROR [org.ovirt.engine.core.ServletUtils] (ajp--127.0.0.1-8702-9) Can't read file /usr/share/ovirt-engine/docs/DocumentationPath.csv for request /docs/DocumentationPath.csv, will send a 404 error response. 2012-11-07 21:23:06,285 INFO [org.ovirt.engine.core.vdsbroker.irsbroker.GetVmsInfoVDSCommand] (ajp--127.0.0.1-8702-11) START, GetVmsInfoVDSCommand( storagePoolId = f90a0d1c-06ca-11e2-a05b-00151712f280, ignoreFailoverLimit = false, compatabilityVersion = null, storageDomainId = 1130b87a-3b34-45d6-8016-d435825c68ef, vmIdList = null), log id: 4863ae13 2012-11-07 21:23:06,336 INFO [org.ovirt.engine.core.vdsbroker.irsbroker.GetVmsInfoVDSCommand] (ajp--127.0.0.1-8702-11) FINISH, GetVmsInfoVDSCommand, log id: 4863ae13 2012-11-07 21:23:12,622 INFO [org.ovirt.engine.core.bll.VdsLoadBalancer] (QuartzScheduler_Worker-84) VdsLoadBalancer: Starting load balance for cluster: Horde, algorithm: EvenlyDistribute. 2012-11-07 21:23:12,625 INFO [org.ovirt.engine.core.bll.VdsLoadBalancer] (QuartzScheduler_Worker-84) VdsLoadBalancer: high util: 75, low util: 0, duration: 2, threashold: 80 2012-11-07 21:23:12,695 INFO [org.ovirt.engine.core.bll.VdsLoadBalancingAlgorithm] (QuartzScheduler_Worker-84) VdsLoadBalancer: number of relevant vdss (no migration, no pending): 1. 2012-11-07 21:23:12,698 INFO
[Users] export domain issues with latest nightly/GIT Master
I have noted some export domain issues with builds from the latest GIT Master. 1) When importing a VM, the disks subtab of the VM to import will not show the disks. Instead it shows the blinking progress squares. This is persistent and will it will not show information on the disks to be imported. This stopped working about 2 weeks back. Expected is to show the disks to import as well as set provisioning type and destination storage domain. 2) The import process does does not respect thin provisioning. Importing a VM without checking the collapse snapshots box (there are actually no snapshots to collapse...) results in the imported disks being imported thickly provisioned. If the collapse snapshots box is checked thin provisioning is respected and the disks import thin provisioned. This may be intertwined with the prior issue. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain
Itamar, I concur with you. I am aware of the dangers here (I edited metadata once already to test recovering a storage domain in a test setup). It took me a few tries and looking through the code that generate and manipulates that data to get it right. However perhaps as a feature request it would be prudent to allow for the ability to choose or change which domain is the master without taking down all the other domains. If you have a-lot of domains this really becomes problematic to play musical chairs with the domains to get the one you want to be master elected. This also requires you to take down VM's which makes it even more problematic. Can we get a make this domain the master feature added to the UI/Cli? -DHC On Thu, Nov 1, 2012 at 2:07 AM, Itamar Heim ih...@redhat.com wrote: On 11/01/2012 03:01 AM, Dead Horse wrote: This is good info. Given that the metadata can be edited is there a way to redesignate which storage domain should be the master? I have run into a couple of times in which for various reasons the master domain could not be reached and another storage domain was elected as master. Is there a way to force an election or edit the metadata to redesignate the original master domain back as master again? i heartly suggest to not try and mess with that on your own. you can move other storage domains to maint, to force engine to reconstruct the master on another domain. (again, do *not* do it on your own via metadata) - DHC On Wed, Oct 31, 2012 at 12:43 PM, Itamar Heim ih...@redhat.com mailto:ih...@redhat.com wrote: On 10/31/2012 07:36 PM, Simon Grinberg wrote: - Original Message - From: Itamar Heim ih...@redhat.com mailto: ih...@redhat.com To: Michael Ayers ayer...@redhat.com mailto:ayer...@redhat.com Cc: users@ovirt.org mailto:users@ovirt.org Sent: Tuesday, October 30, 2012 10:57:52 PM Subject: Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain On 10/30/2012 09:29 PM, Michael Ayers wrote: Hey All, I ran into this issue myself where I needed to convert a data domain to an export domain in order to recover virtual guests from a corrupted ovirt/rhevm instance into a new ovirt/rhevm instance. This wasn’t documented anywhere that I saw but with the help of Itamar Heim and an well timed email to the list from Igor Lvovsky last night I was able to do this and wanted to send to the list a documented procedure for how to modify the metadata of the data domain prior to import as an export domain. This procedure works for both RHEV-M and Ovirt. Let me know if you have any questions. glad it worked. Will this work if the recovered storage domain was not the master? no. only for nfs storage domain, only if it was master. care to wikify for others? thanks, Itamar Original Data Domain Metadata File --**__ CLASS=Data DESCRIPTION=vm-storage IOOPTIMEOUTSEC=10 LEASERETRIES=3 LEASETIMESEC=60 LOCKPOLICY= LOCKRENEWALINTERVALSEC=5 MASTER_VERSION=2 POOL_DESCRIPTION=MyPool POOL_DOMAINS=dd8bc200-6e28-__** 4185-bfe0-f0affb94f283:Active,**__ddefdf6c-ef68-419c-9f72-__** 76d27bf7d788:Active,66b3b243-_**_6cc6-465f-b130-6f2cd0b70514:_**_Active POOL_SPM_ID=2 POOL_SPM_LVER=17 POOL_UUID=a207f052-f4bf-44a3-_**_b637-c6d2020a7c41 REMOTE_PATH=nfsserver:/ovirt/_**_vm-storage ROLE=Master SDUUID=66b3b243-6cc6-465f-__**b130-6f2cd0b70514 TYPE=NFS VERSION=0 _SHA_CKSUM=__**009fa538321ac56749669127f43cc7** __54aa59d398 Diff between Original DD Metadata File and ED Metadata File --**__ --- metadata-data-storage 2012-10-30 12:24:52.484006958 -0700 +++ metadata-exp-storage2012-10-30 12:14:59.043807789 -0700 @@ -1,5 +1,5 @@ -CLASS=Data -DESCRIPTION=vm-storage +CLASS=Backup +DESCRIPTION=export-storage
Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain
Even better ;) I thought I saw mention on the mailing lists of some notion of storage domain import/export is that also in the works? On Thu, Nov 1, 2012 at 9:47 AM, Itamar Heim ih...@redhat.com wrote: On 11/01/2012 04:45 PM, Dead Horse wrote: Itamar, I concur with you. I am aware of the dangers here (I edited metadata once already to test recovering a storage domain in a test setup). It took me a few tries and looking through the code that generate and manipulates that data to get it right. However perhaps as a feature request it would be prudent to allow for the ability to choose or change which domain is the master without taking down all the other domains. If you have a-lot of domains this really becomes problematic to play musical chairs with the domains to get the one you want to be master elected. This also requires you to take down VM's which makes it even more problematic. Can we get a make this domain the master feature added to the UI/Cli? the plan is to remove the need for a master domain to begin with (SDM feature) -DHC On Thu, Nov 1, 2012 at 2:07 AM, Itamar Heim ih...@redhat.com mailto:ih...@redhat.com wrote: On 11/01/2012 03:01 AM, Dead Horse wrote: This is good info. Given that the metadata can be edited is there a way to redesignate which storage domain should be the master? I have run into a couple of times in which for various reasons the master domain could not be reached and another storage domain was elected as master. Is there a way to force an election or edit the metadata to redesignate the original master domain back as master again? i heartly suggest to not try and mess with that on your own. you can move other storage domains to maint, to force engine to reconstruct the master on another domain. (again, do *not* do it on your own via metadata) - DHC On Wed, Oct 31, 2012 at 12:43 PM, Itamar Heim ih...@redhat.com mailto:ih...@redhat.com mailto:ih...@redhat.com mailto:ih...@redhat.com wrote: On 10/31/2012 07:36 PM, Simon Grinberg wrote: - Original Message - From: Itamar Heim ih...@redhat.com mailto:ih...@redhat.com mailto:ih...@redhat.com mailto:ih...@redhat.com To: Michael Ayers ayer...@redhat.com mailto:ayer...@redhat.com mailto:ayer...@redhat.com mailto:ayer...@redhat.com Cc: users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org mailto:users@ovirt.org Sent: Tuesday, October 30, 2012 10:57:52 PM Subject: Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain On 10/30/2012 09:29 PM, Michael Ayers wrote: Hey All, I ran into this issue myself where I needed to convert a data domain to an export domain in order to recover virtual guests from a corrupted ovirt/rhevm instance into a new ovirt/rhevm instance. This wasn’t documented anywhere that I saw but with the help of Itamar Heim and an well timed email to the list from Igor Lvovsky last night I was able to do this and wanted to send to the list a documented procedure for how to modify the metadata of the data domain prior to import as an export domain. This procedure works for both RHEV-M and Ovirt. Let me know if you have any questions. glad it worked. Will this work if the recovered storage domain was not the master? no. only for nfs storage domain, only if it was master. care to wikify for others? thanks, Itamar Original Data Domain Metadata File --** CLASS=Data DESCRIPTION=vm-storage IOOPTIMEOUTSEC=10 LEASERETRIES=3 LEASETIMESEC=60 LOCKPOLICY= LOCKRENEWALINTERVALSEC=5
Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain
This is good info. Given that the metadata can be edited is there a way to redesignate which storage domain should be the master? I have run into a couple of times in which for various reasons the master domain could not be reached and another storage domain was elected as master. Is there a way to force an election or edit the metadata to redesignate the original master domain back as master again? - DHC On Wed, Oct 31, 2012 at 12:43 PM, Itamar Heim ih...@redhat.com wrote: On 10/31/2012 07:36 PM, Simon Grinberg wrote: - Original Message - From: Itamar Heim ih...@redhat.com To: Michael Ayers ayer...@redhat.com Cc: users@ovirt.org Sent: Tuesday, October 30, 2012 10:57:52 PM Subject: Re: [Users] Documentation: Storage Domain conversion from Data Domain to Export Domain On 10/30/2012 09:29 PM, Michael Ayers wrote: Hey All, I ran into this issue myself where I needed to convert a data domain to an export domain in order to recover virtual guests from a corrupted ovirt/rhevm instance into a new ovirt/rhevm instance. This wasn’t documented anywhere that I saw but with the help of Itamar Heim and an well timed email to the list from Igor Lvovsky last night I was able to do this and wanted to send to the list a documented procedure for how to modify the metadata of the data domain prior to import as an export domain. This procedure works for both RHEV-M and Ovirt. Let me know if you have any questions. glad it worked. Will this work if the recovered storage domain was not the master? no. only for nfs storage domain, only if it was master. care to wikify for others? thanks, Itamar Original Data Domain Metadata File --** CLASS=Data DESCRIPTION=vm-storage IOOPTIMEOUTSEC=10 LEASERETRIES=3 LEASETIMESEC=60 LOCKPOLICY= LOCKRENEWALINTERVALSEC=5 MASTER_VERSION=2 POOL_DESCRIPTION=MyPool POOL_DOMAINS=dd8bc200-6e28-**4185-bfe0-f0affb94f283:Active,** ddefdf6c-ef68-419c-9f72-**76d27bf7d788:Active,66b3b243-** 6cc6-465f-b130-6f2cd0b70514:**Active POOL_SPM_ID=2 POOL_SPM_LVER=17 POOL_UUID=a207f052-f4bf-44a3-**b637-c6d2020a7c41 REMOTE_PATH=nfsserver:/ovirt/**vm-storage ROLE=Master SDUUID=66b3b243-6cc6-465f-**b130-6f2cd0b70514 TYPE=NFS VERSION=0 _SHA_CKSUM=**009fa538321ac56749669127f43cc7**54aa59d398 Diff between Original DD Metadata File and ED Metadata File --** --- metadata-data-storage 2012-10-30 12:24:52.484006958 -0700 +++ metadata-exp-storage2012-10-30 12:14:59.043807789 -0700 @@ -1,5 +1,5 @@ -CLASS=Data -DESCRIPTION=vm-storage +CLASS=Backup +DESCRIPTION=export-storage IOOPTIMEOUTSEC=10 LEASERETRIES=3 LEASETIMESEC=60 @@ -7,13 +7,12 @@ LOCKRENEWALINTERVALSEC=5 MASTER_VERSION=2 POOL_DESCRIPTION=MyPool -POOL_DOMAINS=dd8bc200-6e28-**4185-bfe0-f0affb94f283:Active,** ddefdf6c-ef68-419c-9f72-**76d27bf7d788:Active,66b3b243-** 6cc6-465f-b130-6f2cd0b70514:**Active +POOL_DOMAINS= POOL_SPM_ID=2 POOL_SPM_LVER=17 -POOL_UUID=a207f052-f4bf-44a3-**b637-c6d2020a7c41 -REMOTE_PATH=nfsserver:/ovirt/**vm-storage -ROLE=Master +POOL_UUID= +REMOTE_PATH=nfsserver:/ovirt/**export-storage +ROLE=Regular SDUUID=66b3b243-6cc6-465f-**b130-6f2cd0b70514 TYPE=NFS VERSION=0 -_SHA_CKSUM=**009fa538321ac56749669127f43cc7**54aa59d398 Thanks, Michael --**--- Michael J. Ayers Red Hat Inc. Solutions Architect e: ayer...@redhat.com mailto:ayer...@redhat.com w: www.redhat.com http://www.redhat.com/ __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] Quota Memory Max in UI
I noticed that (using latest GIT Master) the Quota UI will only allow entering of an integer number 1 - 65535 into the Memory limit to quota field. Thus in MB this limits max memory per quota to 64GB. I am assuming this to be a bug as obviously one would want to allow for much more than 64GB. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] Quota Memory Max in UI
Bug 870213 was opened regarding this issue. -DHC On Thu, Oct 25, 2012 at 2:39 PM, Greg Padgett gpadg...@redhat.com wrote: On 10/25/2012 02:14 PM, Dead Horse wrote: I noticed that (using latest GIT Master) the Quota UI will only allow entering of an integer number 1 - 65535 into the Memory limit to quota field. Thus in MB this limits max memory per quota to 64GB. I am assuming this to be a bug as obviously one would want to allow for much more than 64GB. - DHC __**_ Users mailing list Users@ovirt.org http://lists.ovirt.org/**mailman/listinfo/usershttp://lists.ovirt.org/mailman/listinfo/users Thanks for bringing this up. I see it, too, and don't know why there would be such a limit. Would you mind opening up a bug to help track the issue? Thanks, Greg ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] [Spice-devel] mouse problem with muiltiple monitors (was HowTo: Spice ActiveX Plugin/Virt Viewer Console on oVirt 3.1)
I've actually started to notice that this occurs natively as well (I rarely if ever use the activex console on a native windows install). However I have noted that with a Native Windows install and the Activex Remote Viewer version 0.5.3 that alot of guests have mouse issues w/o spice vdagent or the windows spice agent in place (very annoying during LFC guests with no knowledge of spice mice). Windows installs can be slipstreamed with the spice drivers to fix this but thats really a pain. Linux guests however not as easy. This actually seems to a repeat of this issue: http://lists.ovirt.org/pipermail/users/2012-February/000351.html (for which a fix was applied to windows spicec + activex NOT remote-viewer) possible regression? Using the usbtablet custom vdsm hook does indeed solve issues with guests that encounter this issue. I have noted it within older linux guests (no surprise they know nothing of a spice mouse), windows guests, solaris guests, some of the older opensuse and ubuntu as well as fedora. The most confusing one was RHEL 6.x which without the spice vdagent loaded will experience the cursor jumping in and out of the window or randomly on the spice display. I note the one downside of the usbtablet is that power users are able to view custom hooks in the PUP UI, but cannot set them as only the admin role not user role can do this. That is to say the power user can look at and even set the hook but is stopped short of committing it to the VM settings. - DHC On Fri, Oct 5, 2012 at 5:18 AM, Christophe Fergeau cferg...@redhat.comwrote: On Thu, Sep 13, 2012 at 12:26:43PM +0300, Itamar Heim wrote: On 09/13/2012 10:24 AM, Karli Sjöberg wrote: 13 sep 2012 kl. 01.21 skrev Dead Horse: Thank you! glad to be able to help ;) As Itamar mentioned if you are running the spice client inside a VM (dunno if this is case) you will need the guest paravirtual driver and/or services for mouse handling. I have observed exactly this behavior before when running the spice client in a VM when the guest tools/drivers for mouse handling are not present (In my case most of the time VirtualBox). Very amusing bug:) But cripples SPICE´s usage. I have tested this from 5 different physical machines running Win7/IE, and this behavior shows itself only on machines with more than one monitor, or a laptop with another monitor attached .e.g. These guest tools you both mention, would these be http://spice-space.org/download/binaries/spice-guest-tools-0.1.exe;? And can you install them in a physical machine as well? cc-ing spice-devel to see if they have insights on your issue Could this be related to https://bugzilla.redhat.com/show_bug.cgi?id=852841 ? Christophe ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] vdsm/engine do not like Infiniband
I have updated the bug. The only side effects I noted from this is that VDSM does not seem to be able to read/report TX/RX stats IB cards properly. Also the MAC Address fields in the admin/user portals seem to be of fixed size and the IB card HW address overflows the fields. - DHC On Thu, Sep 20, 2012 at 3:14 AM, Dan Kenigsberg dan...@redhat.com wrote: On Fri, Sep 14, 2012 at 02:13:37PM -0500, Dead Horse wrote: This is a test setup so no worries about future breakage via upgrade. I ended up stopping the engine service, dumping the database and altering the the table vds_interface -- column mac_addr and increasing the char varying length from 20 to 60. I then restore the altered database and go about business as usual. Please note in the BZ that this is the only change that is required. It would make pushing this upstream much easier. Thanks! I had to make the edit offline because there are quite a few DB views and rules dependent on that table. - DHC On Fri, Sep 14, 2012 at 2:51 AM, Itamar Heim ih...@redhat.com wrote: On 09/14/2012 06:59 AM, Dead Horse wrote: Bug opened BZ857294 ( https://bugzilla.redhat.com/**show_bug.cgi?id=857294 https://bugzilla.redhat.com/show_bug.cgi?id=857294 ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] vdsm/engine do not like Infiniband
This is a test setup so no worries about future breakage via upgrade.
I ended up stopping the engine service, dumping the database and altering
the the table vds_interface -- column mac_addr and increasing the char
varying length from 20 to 60.
I then restore the altered database and go about business as usual.
I had to make the edit offline because there are quite a few DB views and
rules dependent on that table.
- DHC
On Fri, Sep 14, 2012 at 2:51 AM, Itamar Heim ih...@redhat.com wrote:
On 09/14/2012 06:59 AM, Dead Horse wrote:
Bug opened BZ857294
(https://bugzilla.redhat.com/**show_bug.cgi?id=857294https://bugzilla.redhat.com/show_bug.cgi?id=857294
)
What would be a suggestion to work around this ATM and still be able to
use and keep the IB cards in the system?
a bit risky wrt future upgrade which you'll need to carefully test (and
better if you get it first changed upstream...)
change the field to be lengthier (i remember the discussion around this,
but google'd a bit and couldn't find it)
- DHC
On Thu, Sep 13, 2012 at 6:49 PM, Andrew Cathrow acath...@redhat.com
mailto:acath...@redhat.com wrote:
Yes you are right there is a known issue with the mac address length
in IB cards. IIRC there was a hardware vendor who was working on a
patch to extend the length and add the appropriate validation.
Dead Horse deadhorseconsult...@gmail.com
mailto:deadhorseconsulting@**gmail.comdeadhorseconsult...@gmail.com
wrote:
When attempting to activate/add a host with Infiniband cards present:
vdsm.log
Thread-14::DEBUG::2012-09-13
16:16:29,306::BindingXMLRPC::**884::vds::(wrapper) return
getCapabilities
with {'status': {'message': 'Done', 'code': 0}, 'info':
{'HBAInventory':
{'iSCSI': [{'InitiatorName': 'iqn.2012-09.net.azeroth:**kezan'}],
'FC': []},
'packages2': {'kernel': {'release': '200.29.3.el6uek.x86_64',
'buildtime':
1346177011.0, 'version': '2.6.39'}, 'spice-server': {'release':
'10.el6',
'buildtime': 1340343617L, 'version': '0.10.1'}, 'vdsm': {'release':
'435.el6', 'buildtime': 1347508928L, 'version': '4.10.0'}, 'qemu-kvm':
{'release': '2.295.el6_3.2', 'buildtime': 1346867328L, 'version':
'0.12.1.2'}, 'libvirt': {'release': '21.el6_3.4', 'buildtime':
1345740034L,
'version': '0.9.10'}, 'qemu-img': {'release': '2.295.el6_3.2',
'buildtime':
1346867328L, 'version': '0.12.1.2'}}, 'cpuModel': 'Intel(R) Xeon(R)
CPU E5540 @ 2.53GHz', 'hooks': {}, 'vmTypes': ['kvm'],
'supportedProtocols': ['2.2', '2.3'], 'networks': {'ovirtmgmt':
{'iface':
'ovirtmgmt', 'addr': '3.57.111.31', 'cfg': {'IPV6INIT': 'no',
'IPADDR':
'3.57.111.31', 'DELAY': '0', 'PREFIX': '22', 'BOOTPROTO': 'none',
'STP':
'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge', 'ONBOOT': 'yes'},
'mtu':
'1500', 'netmask': '255.255.252.0', 'stp': 'off', 'bridged': True,
'gateway': '0.0.0.0', 'ports': ['bond2']}, 'maelstrom': {'iface':
'maelstrom', 'addr': '', 'cfg': {'IPV6INIT': 'no', 'DELAY': '0',
'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'maelstrom', 'TYPE':
'Bridge',
'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '', 'stp': 'off',
'bridged':
True, 'gateway': '0.0.0.0', 'ports': ['bond0']}, 'twistingnether':
{'iface': 'twistingnether', 'addr': '3.70.204.31', 'cfg': {'IPV6INIT':
'no', 'IPADDR': '3.70.204.31', 'DELAY': '0', 'PREFIX': '22',
'BOOTPROTO':
'none', 'STP': 'off', 'DEVICE': 'twistingnether', 'TYPE': 'Bridge',
'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '255.255.252.0', 'stp':
'off',
'bridged': True, 'gateway': '3.70.204.254', 'ports': ['bond1']}},
'bridges': {'ovirtmgmt': {'addr': '3.57.111.31', 'cfg': {'IPV6INIT':
'no',
'IPADDR': '3.57.111.31', 'DELAY': '0', 'PREFIX': '22', 'BOOTPROTO':
'none',
'STP': 'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge', 'ONBOOT':
'yes'},
'mtu': '1500', 'netmask': '255.255.252.0', 'stp': 'off', 'ports':
['bond2']}, 'maelstrom': {'addr': '', 'cfg': {'IPV6INIT': 'no',
'DELAY':
'0', 'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'maelstrom', 'TYPE':
'Bridge', 'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '', 'stp':
'off',
'ports': ['bond0']}, 'twistingnether': {'addr': '3.70.204.31', 'cfg':
{'IPV6INIT': 'no', 'IPADDR': '3.70.204.31', 'DELAY': '0', 'PREFIX':
'22',
'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'twistingnether', 'TYPE':
'Bridge', 'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '255.255.252.0',
'stp': 'off', 'ports': ['bond1']}}, 'uuid':
'FF28----**00212883CE4C_00:15:17:00:10:**0a',
'lastClientIface': 'twistingnether', 'nics': {'ib0': {'addr':
'192.168.1.1', 'cfg': {'IPV6INIT': 'no', 'IPADDR': '192.168.1.1',
'MTU':
'65520', 'PREFIX': '16', 'BOOTPROTO': 'none', 'CONNECTED_MODE': 'yes',
'DEVICE': 'ib0', 'TYPE': 'Infiniband', 'ONBOOT': 'yes'}, 'mtu
Re: [Users] vdsm/engine do not like Infiniband
Bug opened BZ857294 (https://bugzilla.redhat.com/show_bug.cgi?id=857294)
What would be a suggestion to work around this ATM and still be able to use
and keep the IB cards in the system?
- DHC
On Thu, Sep 13, 2012 at 6:49 PM, Andrew Cathrow acath...@redhat.com wrote:
Yes you are right there is a known issue with the mac address length in IB
cards. IIRC there was a hardware vendor who was working on a patch to
extend the length and add the appropriate validation.
Dead Horse deadhorseconsult...@gmail.com wrote:
When attempting to activate/add a host with Infiniband cards present:
vdsm.log
Thread-14::DEBUG::2012-09-13
16:16:29,306::BindingXMLRPC::884::vds::(wrapper) return getCapabilities
with {'status': {'message': 'Done', 'code': 0}, 'info': {'HBAInventory':
{'iSCSI': [{'InitiatorName': 'iqn.2012-09.net.azeroth:kezan'}], 'FC': []},
'packages2': {'kernel': {'release': '200.29.3.el6uek.x86_64', 'buildtime':
1346177011.0, 'version': '2.6.39'}, 'spice-server': {'release': '10.el6',
'buildtime': 1340343617L, 'version': '0.10.1'}, 'vdsm': {'release':
'435.el6', 'buildtime': 1347508928L, 'version': '4.10.0'}, 'qemu-kvm':
{'release': '2.295.el6_3.2', 'buildtime': 1346867328L, 'version':
'0.12.1.2'}, 'libvirt': {'release': '21.el6_3.4', 'buildtime': 1345740034L,
'version': '0.9.10'}, 'qemu-img': {'release': '2.295.el6_3.2', 'buildtime':
1346867328L, 'version': '0.12.1.2'}}, 'cpuModel': 'Intel(R) Xeon(R)
CPU E5540 @ 2.53GHz', 'hooks': {}, 'vmTypes': ['kvm'],
'supportedProtocols': ['2.2', '2.3'], 'networks': {'ovirtmgmt': {'iface':
'ovirtmgmt', 'addr': '3.57.111.31', 'cfg': {'IPV6INIT': 'no', 'IPADDR':
'3.57.111.31', 'DELAY': '0', 'PREFIX': '22', 'BOOTPROTO': 'none', 'STP':
'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge', 'ONBOOT': 'yes'}, 'mtu':
'1500', 'netmask': '255.255.252.0', 'stp': 'off', 'bridged': True,
'gateway': '0.0.0.0', 'ports': ['bond2']}, 'maelstrom': {'iface':
'maelstrom', 'addr': '', 'cfg': {'IPV6INIT': 'no', 'DELAY': '0',
'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'maelstrom', 'TYPE': 'Bridge',
'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '', 'stp': 'off', 'bridged':
True, 'gateway': '0.0.0.0', 'ports': ['bond0']}, 'twistingnether':
{'iface': 'twistingnether', 'addr': '3.70.204.31', 'cfg': {'IPV6INIT':
'no', 'IPADDR': '3.70.204.31', 'DELAY': '0', 'PREFIX': '22', 'BOOTPROTO':
'none', 'STP': 'off', 'DEVICE': 'twistingnether', 'TYPE': 'Bridge',
'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '255.255.252.0', 'stp': 'off',
'bridged': True, 'gateway': '3.70.204.254', 'ports': ['bond1']}},
'bridges': {'ovirtmgmt': {'addr': '3.57.111.31', 'cfg': {'IPV6INIT': 'no',
'IPADDR': '3.57.111.31', 'DELAY': '0', 'PREFIX': '22', 'BOOTPROTO': 'none',
'STP': 'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge', 'ONBOOT': 'yes'},
'mtu': '1500', 'netmask': '255.255.252.0', 'stp': 'off', 'ports':
['bond2']}, 'maelstrom': {'addr': '', 'cfg': {'IPV6INIT': 'no', 'DELAY':
'0', 'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'maelstrom', 'TYPE':
'Bridge', 'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '', 'stp': 'off',
'ports': ['bond0']}, 'twistingnether': {'addr': '3.70.204.31', 'cfg':
{'IPV6INIT': 'no', 'IPADDR': '3.70.204.31', 'DELAY': '0', 'PREFIX': '22',
'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'twistingnether', 'TYPE':
'Bridge', 'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '255.255.252.0',
'stp': 'off', 'ports': ['bond1']}}, 'uuid':
'FF28----00212883CE4C_00:15:17:00:10:0a',
'lastClientIface': 'twistingnether', 'nics': {'ib0': {'addr':
'192.168.1.1', 'cfg': {'IPV6INIT': 'no', 'IPADDR': '192.168.1.1', 'MTU':
'65520', 'PREFIX': '16', 'BOOTPROTO': 'none', 'CONNECTED_MODE': 'yes',
'DEVICE': 'ib0', 'TYPE': 'Infiniband', 'ONBOOT': 'yes'}, 'mtu': '65520',
'netmask': '255.255.0.0', 'hwaddr':
'80:00:04:04:fe:80:00:00:00:00:00:00:00:12:79:ff:ff:b5:b0:f1', 'speed': 0},
'ib1': {'addr': '', 'cfg': {'IPV6INIT': 'no', 'MTU': '65520', 'BOOTPROTO':
'none', 'CONNECTED_MODE': 'yes', 'DEVICE': 'ib1', 'TYPE': 'Infiniband',
'ONBOOT': 'no'}, 'mtu': '2044', 'netmask': '', 'hwaddr':
'80:00:04:05:fe:80:00:00:00:00:00:00:00:12:79:ff:ff:b5:b0:f2', 'speed': 0},
'eth5': {'permhwaddr': '00:15:17:00:10:0B', 'addr': '', 'cfg': {'SLAVE':
'yes', 'IPV6INIT': 'no', 'HWADDR': '00:15:17:00:10:0B', 'BOOTPROTO':
'none', 'MASTER': 'bond2', 'DEVICE': 'eth5', 'ONBOOT': 'yes'}, 'mtu':
'1500', 'netmask': '', 'hwaddr': '00:15:17:00:10:0a', 'speed': 1000},
'eth4': {'permhwaddr': '00:15:17:00:10:0A', 'addr': '', 'cfg': {'SLAVE':
'yes', 'IPV6INIT': 'no', 'HWADDR': '00:15:17:00:10:0A', 'BOOTPROTO':
'none', 'MASTER': 'bond2', 'DEVICE': 'eth4', 'ONBOOT': 'yes'}, 'mtu':
'1500', 'netmask': '', 'hwaddr': '00:15:17:00:10:0a', 'speed': 1000},
'eth3': {'permhwaddr': '00:21:28:83:CE:4B', 'addr': '', 'cfg': {'SLAVE':
'yes', 'HWADDR': '00:21:28:83:CE:4B', 'BOOTPROTO': 'none', 'MASTER':
'bond1', 'DEVICE': 'eth3', 'ONBOOT': 'yes'}, 'mtu': '1500', 'netmask': '',
'hwaddr': '00:21:28:83:ce:4a
Re: [Users] HowTo: Spice ActiveX Plugin/Virt Viewer Console on oVirt 3.1
Thank you! glad to be able to help ;) As Itamar mentioned if you are running the spice client inside a VM (dunno if this is case) you will need the guest paravirtual driver and/or services for mouse handling. I have observed exactly this behavior before when running the spice client in a VM when the guest tools/drivers for mouse handling are not present (In my case most of the time VirtualBox). - DHC On Wed, Sep 12, 2012 at 3:49 PM, Itamar Heim ih...@redhat.com wrote: On 09/12/2012 05:21 PM, Karli Sjöberg wrote: This is awesome work, thank you so much for this! For our users and administrators, having a button you just click to get a console is sooo much easier than having to copy-paste a bunch of cryptic addresses and passwords, and having to use the infamous, dangerous hacker tool VNC:) I followed the wiki instructions and a vital piece missing is to restart ovirt-engine before trying to install the spice-client from the install portal. I went back and read through the email you first posted and the restart was mentioned there so I did and then it worked! Wohoo! So you might want to add that point to the wiki page as well, to avoid future confusion. But...when I have virt-viewer(-0.5.3) installed, log in to webadmin and click for console on a guest, the console shows up. But when I try to go into the console window with the mouse cursor, the cursor jumps off to one of the edges of my *monitor*, and not into the console window- as you would expect. Maybe virt-viewer mistakenly thinks that the edges of it´s window really are the edges of the whole monitor? Am I holding it wrong? do you have spice guest side drivers for local mouse handling? /Karli 6 sep 2012 kl. 22.36 skrev Dead Horse: Wiki page added. Feel free to change/edit/contribute or provide feedback. Link: http://wiki.ovirt.org/wiki/**How_to_Connect_to_SPICE_** Console_With_Portalhttp://wiki.ovirt.org/wiki/How_to_Connect_to_SPICE_Console_With_Portal - DHC On Thu, Aug 30, 2012 at 11:05 PM, Robert Middleswarth rob...@middleswarth.net mailto:robert@middleswarth.**netrob...@middleswarth.net wrote: You should be able to create an ID. Once you create it several people can update your account I am one of them. Thanks Robert On 08/30/2012 11:17 PM, Dead Horse wrote: I would be happy to wiki'fy it ;) I have requested a wiki account so I will just need someone to grant the proper create/edit page access or create a stub page and allow edit access to it. - DHC On Thu, Aug 30, 2012 at 3:03 PM, Douglas Landgraf dougsl...@redhat.com mailto:dougsl...@redhat.com wrote: On 08/30/2012 02:22 AM, Yaniv Kaul wrote: Thanks! Can you wiki'fy it (@ wiki.ovirt.org http://wiki.ovirt.org/) ? Y. +1 Thanks! --**-- ** I have seen this question asked many times on this list and the spice-devel list. Now having figured out how to make it work I will provide the answer to the rest of the community. *NOTE* this only applies to Windows/Internet Explorer users. There exists no other option other than the html5 spice console (still beta/in development) for windows users ATM. This will also only work with Internet Explorer. Basic Steps: - Install an oVirt server. - Add at least one node - Setup storage/iso/export domains Advanced steps - Start by downloading: http://elmarco.fedorapeople.**org/spice.cabhttp://elmarco.fedorapeople.org/spice.cab - Create a directory for it oVirt looks by default in /usr/share/spice - Rename spice.cab to SpiceX.cab and copy it into /usr/share/spice - Now edit /usr/share/ovirt-engine/**engine.ear/root.war/WEB-INF and add the following: !-- SpiceX.cab -- servlet servlet-nameSpiceX.cab/**servlet-name servlet-classorg.ovirt.**engine.core.FileServlet/** servlet-class init-param param-nametype/param-name param-valueapplication/** octet-stream/param-value /init-param init-param param-namefile/param-name param-value/usr/share/spice/** SpiceX.cab/param-value /init-param /servlet servlet-mapping servlet-nameSpiceX.cab/**servlet-name url-pattern/spice/SpiceX.**cab/url-pattern /servlet-mapping - Next create an html file within /usr/share/ovirt-engine/**engine.ear/root.war - In the example below an html file
[Users] HowTo: Using EL6 hosts with Ovirt
I took the liberty of creating a wiki page on how to use EL6 based hosts with the ovirt-engine. It can be found here: http://wiki.ovirt.org/wiki/Using_EL6_hosts_with_Ovirt Feel free to change/edit/contribute or provide feedback. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] EL 6.3+ nodes and 3.1 compatibility?
Rebuilding vdsm - ovirt-3.1 from GIT and according dependent packages did
the trick to get the node added into the 3.1 cluster (EG: logrotate,
gluster, etc). Next I ran into the issue with the current RHEL 6.3 sanlock
not starting due to wdmd failing startup due to a watchdog not being loaded
(https://bugzilla.redhat.com/show_bug.cgi?id=832935). Fixed that by
building and installing a newer sanlock where this was fixed.
- DHC
On Sat, Aug 11, 2012 at 4:31 PM, Itamar Heim ih...@redhat.com wrote:
On 08/10/2012 04:21 PM, Dead Horse wrote:
When I try adding an EL 6.3 based node to an ovirt 3.1 cluster, ovirt
complains that the host must be run in 3.0 compatibility mode and then
bombs out.
you should create a 3.0 dc/cluster for this version of vdsm.
RHEL does not yet have a VDSM version with 3.1 support.
The version of VDSM on the node is:
vdsm-4.9-113.1.el6.x86_64
vdsm-cli-4.9-113.1.el6.x86_64
Other misc versions:
kernel-2.6.32-279.2.1.el6.x86_
64
libvirt-0.9.10-21.el6_3.3.x86_**64
libvirt-python-0.9.10-21.el6_**3.3.x86_64
qemu-img-0.12.1.2-2.295.el6_3.**1.x86_64
qemu-kvm-0.12.1.2-2.295.el6_3.**1.x86_64
gpxe-roms-qemu-0.9.7-6.9.el6.**noarch
qemu-kvm-tools-0.12.1.2-2.295.**el6_3.1.x86_64
spice-server-0.10.1-10.el6.**x86_64
'vdsClient -s 0 getVdsCaps' is does not seem to work since I have SSL
disabled for vdsm so here is the output from the vdsm.log:
Thread-13::DEBUG::2012-08-10 00:46:26,202::clientIF::59::**vds::(wrapper)
return getVdsCapabilities with {'status': {'message': 'Done', 'code':
0}, 'info': {'HBAInventory': {'iSCSI': [{'InitiatorName':
'iqn.1994-05.com.redhat:**decf7eb6ff3f'}], 'FC': []}, 'packages2':
{'kernel': {'release': '279.2.1.el6.x86_64', 'buildtime': 1342732135.0,
'version': '2.6.32'}, 'spice-server': {'release': '10.el6', 'buildtime':
'1340343617', 'version': '0.10.1'}, 'vdsm': {'release': '113.1.el6',
'buildtime': '1344576418', 'version': '4.9'}, 'qemu-kvm': {'release':
'2.295.el6_3.1', 'buildtime': '1343643134', 'version': '0.12.1.2'},
'libvirt': {'release': '21.el6_3.3', 'buildtime': '1342618149',
'version': '0.9.10'}, 'qemu-img': {'release': '2.295.el6_3.1',
'buildtime': '1343643134', 'version': '0.12.1.2'}}, 'cpuModel':
'Intel(R) Xeon(R) CPU E5540 @ 2.53GHz', 'hooks': {},
'vmTypes': ['kvm'], 'supportedProtocols': ['2.2', '2.3'], 'networks':
{'ovirtmgmt': {'addr': '192.168.1.22', 'cfg': {'IPV6INIT': 'no',
'IPADDR': '192.168.1.22', 'DELAY': '0', 'PREFIX': '24', 'BOOTPROTO':
'none', 'STP': 'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge', 'ONBOOT':
'yes'}, 'ports': ['eth0'], 'netmask': '255.255.255.0', 'stp': 'off',
'gateway': '192.168.1.1'}, 'virbr0': {'addr': '192.168.122.1', 'cfg':
{}, 'ports': ['virbr0-nic'], 'netmask': '255.255.255.0', 'stp': 'on',
'gateway': '0.0.0.0'}}, 'uuid':
'FF28----**00212883B51A_00:21:28:83:B5:**16', 'nics':
{'eth3': {'hwaddr': '00:21:28:83:B5:19', 'netmask': '', 'speed': 0,
'addr': ''}, 'eth2': {'hwaddr': '00:21:28:83:B5:18', 'netmask': '',
'speed': 0, 'addr': ''}, 'eth1': {'hwaddr': '00:21:28:83:B5:17',
'netmask': '', 'speed': 0, 'addr': ''}, 'eth0': {'hwaddr':
'00:21:28:83:B5:16', 'netmask': '', 'speed': 1000, 'addr': ''}},
'software_revision': '113.1', 'management_ip': '', 'clusterLevels':
['3.0'], 'cpuFlags':
u'fpu,vme,de,pse,tsc,msr,pae,**mce,cx8,apic,mtrr,pge,mca,**
cmov,pat,pse36,clflush,dts,**acpi,mmx,fxsr,sse,sse2,ss,ht,**
tm,pbe,syscall,nx,rdtscp,lm,**constant_tsc,arch_perfmon,**
pebs,bts,rep_good,xtopology,**nonstop_tsc,aperfmperf,pni,**
dtes64,monitor,ds_cpl,vmx,est,**tm2,ssse3,cx16,xtpr,pdcm,dca,**
sse4_1,sse4_2,popcnt,lahf_lm,**ida,dts,tpr_shadow,vnmi,**
flexpriority,ept,vpid,model_**486,model_pentium,model_**
pentium2,model_pentium3,model_**pentiumpro,model_qemu32,model_**
cpu64-rhel5,model_cpu64-rhel6,**model_kvm32,model_coreduo,**
model_kvm64,model_core2duo,**model_n270,model_Conroe,model_**
Penryn,model_Nehalem,model_**Opteron_G1',
'ISCSIInitiatorName': 'iqn.1994-05.com.redhat:**decf7eb6ff3f', 'memSize':
'36140', 'reservedMem': '256', 'bondings': {'bond4': {'hwaddr':
'00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr': '', 'slaves':
[]}, 'bond0': {'hwaddr': '00:00:00:00:00:00', 'cfg': {}, 'netmask': '',
'addr': '', 'slaves': []}, 'bond1': {'hwaddr': '00:00:00:00:00:00',
'cfg': {}, 'netmask': '', 'addr': '', 'slaves': []}, 'bond2': {'hwaddr':
'00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr': '', 'slaves':
[]}, 'bond3': {'hwaddr': '00:00:00:00:00:00', 'cfg': {}, 'netmask': '',
'addr': '', 'slaves': []}}, 'software_version': '4.9', 'cpuSpeed':
'1600.000', 'cpuSockets': '2', 'vlans': {}, 'cpuCores': '8',
'kvmEnabled': 'true', 'guestOverhead': '65', 'supportedRHEVMs': ['3.0'],
'version_name': 'Snow Man', 'emulatedMachines': [u'rhel6.3.0', u'pc',
u'rhel6.2.0', u'rhel6.1.0', u'rhel6.0.0', u'rhel5.5.0', u'rhel5.4.4',
u'rhel5.4.0'], 'operatingSystem': {'release': '1', 'version': '6.3',
'name': 'RHEL'}}}
Thread-14::DEBUG::2012-08-10
[Users] EL 6.3+ nodes and 3.1 compatibility?
With ovirt 3.1 EL 6.3 nodes will only work with a data center set to 3.0 compatibility mode. Is there any reason(s) why EL 6.3+ nodes would not be allowed in a 3.1 compatibility level cluster? This assumes non use of gluster since adding vdsm-gluster to an EL 6 node requires some work package update/version wise (not impossible). - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] EL 6.3+ nodes and 3.1 compatibility?
When I try adding an EL 6.3 based node to an ovirt 3.1 cluster, ovirt
complains that the host must be run in 3.0 compatibility mode and then
bombs out.
The version of VDSM on the node is:
vdsm-4.9-113.1.el6.x86_64
vdsm-cli-4.9-113.1.el6.x86_64
Other misc versions:
kernel-2.6.32-279.2.1.el6.x86_
64
libvirt-0.9.10-21.el6_3.3.x86_64
libvirt-python-0.9.10-21.el6_3.3.x86_64
qemu-img-0.12.1.2-2.295.el6_3.1.x86_64
qemu-kvm-0.12.1.2-2.295.el6_3.1.x86_64
gpxe-roms-qemu-0.9.7-6.9.el6.noarch
qemu-kvm-tools-0.12.1.2-2.295.el6_3.1.x86_64
spice-server-0.10.1-10.el6.x86_64
'vdsClient -s 0 getVdsCaps' is does not seem to work since I have SSL
disabled for vdsm so here is the output from the vdsm.log:
Thread-13::DEBUG::2012-08-10 00:46:26,202::clientIF::59::vds::(wrapper)
return getVdsCapabilities with {'status': {'message': 'Done', 'code': 0},
'info': {'HBAInventory': {'iSCSI': [{'InitiatorName':
'iqn.1994-05.com.redhat:decf7eb6ff3f'}], 'FC': []}, 'packages2': {'kernel':
{'release': '279.2.1.el6.x86_64', 'buildtime': 1342732135.0, 'version':
'2.6.32'}, 'spice-server': {'release': '10.el6', 'buildtime': '1340343617',
'version': '0.10.1'}, 'vdsm': {'release': '113.1.el6', 'buildtime':
'1344576418', 'version': '4.9'}, 'qemu-kvm': {'release': '2.295.el6_3.1',
'buildtime': '1343643134', 'version': '0.12.1.2'}, 'libvirt': {'release':
'21.el6_3.3', 'buildtime': '1342618149', 'version': '0.9.10'}, 'qemu-img':
{'release': '2.295.el6_3.1', 'buildtime': '1343643134', 'version':
'0.12.1.2'}}, 'cpuModel': 'Intel(R) Xeon(R) CPU E5540 @
2.53GHz', 'hooks': {}, 'vmTypes': ['kvm'], 'supportedProtocols': ['2.2',
'2.3'], 'networks': {'ovirtmgmt': {'addr': '192.168.1.22', 'cfg':
{'IPV6INIT': 'no', 'IPADDR': '192.168.1.22', 'DELAY': '0', 'PREFIX': '24',
'BOOTPROTO': 'none', 'STP': 'off', 'DEVICE': 'ovirtmgmt', 'TYPE': 'Bridge',
'ONBOOT': 'yes'}, 'ports': ['eth0'], 'netmask': '255.255.255.0', 'stp':
'off', 'gateway': '192.168.1.1'}, 'virbr0': {'addr': '192.168.122.1',
'cfg': {}, 'ports': ['virbr0-nic'], 'netmask': '255.255.255.0', 'stp':
'on', 'gateway': '0.0.0.0'}}, 'uuid':
'FF28----00212883B51A_00:21:28:83:B5:16', 'nics': {'eth3':
{'hwaddr': '00:21:28:83:B5:19', 'netmask': '', 'speed': 0, 'addr': ''},
'eth2': {'hwaddr': '00:21:28:83:B5:18', 'netmask': '', 'speed': 0, 'addr':
''}, 'eth1': {'hwaddr': '00:21:28:83:B5:17', 'netmask': '', 'speed': 0,
'addr': ''}, 'eth0': {'hwaddr': '00:21:28:83:B5:16', 'netmask': '',
'speed': 1000, 'addr': ''}}, 'software_revision': '113.1', 'management_ip':
'', 'clusterLevels': ['3.0'], 'cpuFlags':
u'fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,mtrr,pge,mca,cmov,pat,pse36,clflush,dts,acpi,mmx,fxsr,sse,sse2,ss,ht,tm,pbe,syscall,nx,rdtscp,lm,constant_tsc,arch_perfmon,pebs,bts,rep_good,xtopology,nonstop_tsc,aperfmperf,pni,dtes64,monitor,ds_cpl,vmx,est,tm2,ssse3,cx16,xtpr,pdcm,dca,sse4_1,sse4_2,popcnt,lahf_lm,ida,dts,tpr_shadow,vnmi,flexpriority,ept,vpid,model_486,model_pentium,model_pentium2,model_pentium3,model_pentiumpro,model_qemu32,model_cpu64-rhel5,model_cpu64-rhel6,model_kvm32,model_coreduo,model_kvm64,model_core2duo,model_n270,model_Conroe,model_Penryn,model_Nehalem,model_Opteron_G1',
'ISCSIInitiatorName': 'iqn.1994-05.com.redhat:decf7eb6ff3f', 'memSize':
'36140', 'reservedMem': '256', 'bondings': {'bond4': {'hwaddr':
'00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr': '', 'slaves': []},
'bond0': {'hwaddr': '00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr':
'', 'slaves': []}, 'bond1': {'hwaddr': '00:00:00:00:00:00', 'cfg': {},
'netmask': '', 'addr': '', 'slaves': []}, 'bond2': {'hwaddr':
'00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr': '', 'slaves': []},
'bond3': {'hwaddr': '00:00:00:00:00:00', 'cfg': {}, 'netmask': '', 'addr':
'', 'slaves': []}}, 'software_version': '4.9', 'cpuSpeed': '1600.000',
'cpuSockets': '2', 'vlans': {}, 'cpuCores': '8', 'kvmEnabled': 'true',
'guestOverhead': '65', 'supportedRHEVMs': ['3.0'], 'version_name': 'Snow
Man', 'emulatedMachines': [u'rhel6.3.0', u'pc', u'rhel6.2.0', u'rhel6.1.0',
u'rhel6.0.0', u'rhel5.5.0', u'rhel5.4.4', u'rhel5.4.0'], 'operatingSystem':
{'release': '1', 'version': '6.3', 'name': 'RHEL'}}}
Thread-14::DEBUG::2012-08-10 00:46:26,335::clientIF::54::vds::(wrapper)
[192.168.1.20]::call getVdsCapabilities with () {}
Thread-14::DEBUG::2012-08-10
00:46:26,353::utils::579::Storage.Misc.excCmd::(execCmd) '/bin/rpm -q --qf
%{NAME}\t%{VERSION}\t%{RELEASE}\t%{BUILDTIME}\n qemu-kvm' (cwd None)
- DHC
On Fri, Aug 10, 2012 at 6:41 AM, Itamar Heim ih...@redhat.com wrote:
On 08/10/2012 09:16 AM, Michal Skrivanek wrote:
On Aug 10, 2012, at 08:02 , Dead Horse wrote:
With ovirt 3.1 EL 6.3 nodes will only work with a data center set to 3.0
compatibility mode. Is there any reason(s) why EL 6.3+ nodes would not be
allowed in a 3.1 compatibility level cluster?
Why would they only work in 3.0 compatibility? 3.1 nodes work in 3.1
cluster level because, well, they are the same version.
Or I didn't get the question and please
Re: [Users] EL 6.3+ nodes and 3.1 compatibility?
I am not using the dreyou repo. Rather a pure EL 6.3 and it's associated packages. - DHC On Fri, Aug 10, 2012 at 8:07 AM, Robert Middleswarth rob...@middleswarth.net wrote: On 08/10/2012 02:02 AM, Dead Horse wrote: With ovirt 3.1 EL 6.3 nodes will only work with a data center set to 3.0 compatibility mode. Is there any reason(s) why EL 6.3+ nodes would not be allowed in a 3.1 compatibility level cluster? This assumes non use of gluster since adding vdsm-gluster to an EL 6 node requires some work package update/version wise (not impossible). - DHC ___ Users mailing listUsers@ovirt.orghttp://lists.ovirt.org/mailman/listinfo/users What repo are you using? What build number? I have used both gluster and 3.1 compatibily nodes on top of CentOS 6.3. http://www.middleswarth.net/content/installing-ovirt-31-and-glusterfs-using-either-nfs-or-posix-native-file-system -- Thanks Robert Middleswarth @rmiddle (twitter/IRC) ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] EL 6.3+ nodes and 3.1 compatibility?
This is a pure RHEL 6.3 node. Given that I gather in order to make this work I need to build the newer vdsm 4.10.x and update/build the other according dependent packages? - DHC On Fri, Aug 10, 2012 at 8:23 AM, Andrew Cathrow acath...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Robert Middleswarth rob...@middleswarth.net Cc: users@ovirt.org Sent: Friday, August 10, 2012 9:22:30 AM Subject: Re: [Users] EL 6.3+ nodes and 3.1 compatibility? I am not using the dreyou repo. Rather a pure EL 6.3 and it's associated packages. where are you getting your VDSM packages from - because the 3.1 packages aren't in RHEL or it's clones. - DHC On Fri, Aug 10, 2012 at 8:07 AM, Robert Middleswarth rob...@middleswarth.net wrote: On 08/10/2012 02:02 AM, Dead Horse wrote: With ovirt 3.1 EL 6.3 nodes will only work with a data center set to 3.0 compatibility mode. Is there any reason(s) why EL 6.3+ nodes would not be allowed in a 3.1 compatibility level cluster? This assumes non use of gluster since adding vdsm-gluster to an EL 6 node requires some work package update/version wise (not impossible). - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users What repo are you using? What build number? I have used both gluster and 3.1 compatibily nodes on top of CentOS 6.3. http://www.middleswarth.net/content/installing-ovirt-31-and-glusterfs-using-either-nfs-or-posix-native-file-system -- Thanks Robert Middleswarth @rmiddle (twitter/IRC) ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] ovirt-engine upgrade 3.0 to 3.1 issue
Any thoughts here on this, should I file a bug?
Per: http://wiki.ovirt.org/wiki/Second_Release
Working upgrade is one of the release criteria.
I will help in any way possible to get a 3.0 to 3.1 upgrade path working.
- DHC
On Thu, Jul 19, 2012 at 3:40 PM, Dead Horse
deadhorseconsult...@gmail.comwrote:
If it helps here is the contents of upgrade.sh.log:
* QUERY **
create or replace function __temp_fn_create_schema_version ()
returns void
AS $procedure$
begin
if not exists (select 1 from information_schema.tables where table_name =
'schema_version') then
CREATE SEQUENCE schema_version_seq INCREMENT BY 1 START WITH 1;
CREATE TABLE schema_version
(
id INTEGER DEFAULT NEXTVAL('schema_version_seq') NOT NULL,
version varchar(10) NOT NULL,
script varchar(255) NOT NULL,
checksum varchar(128),
installed_by varchar(30) NOT NULL,
started_at timestamp DEFAULT now(),
ended_at timestamp ,
state character varying(15) NOT NULL,
current boolean NOT NULL,
CONSTRAINT schema_version_primary_key PRIMARY KEY (id)
);
insert into
schema_version(version,script,checksum,installed_by,ended_at,state,current)
values
('0300','upgrade/03_00__add_schema_version.sql','0','postgres',now(),'INSTALLED',true);
end if;
END; $procedure$
LANGUAGE plpgsql;
**
CREATE FUNCTION
* QUERY **
select __temp_fn_create_schema_version();
**
* QUERY **
DROP FUNCTION __temp_fn_create_schema_version();
**
DROP FUNCTION
* QUERY **
CREATE OR REPLACE FUNCTION __temp_Upgrade_add_job_table()
RETURNS void
AS $function$
BEGIN
IF NOT EXISTS (SELECT * FROM information_schema.tables WHERE table_name
ILIKE 'job') THEN
-- Add the job table.
CREATE TABLE job
(
job_id UUID NOT NULL,
action_type VARCHAR(50) NOT NULL,
description TEXT NOT NULL,
status VARCHAR(32) NOT NULL,
owner_id UUID,
visible BOOLEAN NOT NULL DEFAULT true,
start_time TIMESTAMP WITH TIME ZONE NOT NULL,
end_time TIMESTAMP WITH TIME ZONE default NULL,
last_update_time TIMESTAMP WITH TIME ZONE default NULL,
correlation_id VARCHAR(50) NOT NULL,
CONSTRAINT pk_jobs PRIMARY KEY(job_id)
)
WITH OIDS;
END IF;
END; $function$
LANGUAGE plpgsql;
**
CREATE FUNCTION
* QUERY **
SELECT * FROM __temp_Upgrade_add_job_table();
**
* QUERY **
DROP FUNCTION __temp_Upgrade_add_job_table();
**
DROP FUNCTION
* QUERY **
insert into
schema_version(version,script,checksum,installed_by,started_at,ended_at,state,current,comment)
values
(trim('03010260'),'upgrade/03_01_0260_add_job_table.sql','c74d309cfb8f9d802e54f2ff66d560b4','postgres',
cast(trim(' 2012-07-19 15:30:52.500231-05') as
timestamp),cast(trim(' 2012-07-19 15:30:52.565532-05') as
timestamp),'INSTALLED',false,'');
**
* QUERY **
create or replace function __temp_fn_create_schema_version ()
returns void
AS $procedure$
begin
if not exists (select 1 from information_schema.tables where table_name =
'schema_version') then
CREATE SEQUENCE schema_version_seq INCREMENT BY 1 START WITH 1;
CREATE TABLE schema_version
(
id INTEGER DEFAULT NEXTVAL('schema_version_seq') NOT NULL,
version varchar(10) NOT NULL,
script varchar(255) NOT NULL,
checksum varchar(128),
installed_by varchar(30) NOT NULL,
started_at timestamp DEFAULT now(),
ended_at timestamp ,
state character varying(15) NOT NULL,
current boolean NOT NULL,
CONSTRAINT schema_version_primary_key PRIMARY KEY (id)
);
insert into
schema_version(version,script,checksum,installed_by,ended_at,state,current)
values
('0300','upgrade/03_00__add_schema_version.sql','0','postgres',now(),'INSTALLED',true);
end if;
END; $procedure$
LANGUAGE plpgsql;
**
CREATE FUNCTION
* QUERY **
select __temp_fn_create_schema_version();
**
* QUERY **
DROP FUNCTION __temp_fn_create_schema_version();
**
DROP FUNCTION
* QUERY **
CREATE OR REPLACE FUNCTION __temp_Upgrade_add_job_table()
RETURNS void
AS $function$
BEGIN
IF NOT EXISTS (SELECT * FROM information_schema.tables WHERE table_name
ILIKE 'job') THEN
-- Add the job table.
CREATE TABLE job
(
job_id UUID NOT NULL,
action_type VARCHAR(50) NOT NULL,
description TEXT NOT NULL,
status VARCHAR(32) NOT NULL,
owner_id UUID,
visible BOOLEAN NOT NULL DEFAULT
Re: [Users] ovirt-engine upgrade 3.0 to 3.1 issue
Eli, I am following the procedure for DB backup/restore from here: http://wiki.ovirt.org/wiki/Backup_engine_db and here: http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Virtualization/3.0/html/Administration_Guide/Administration_Guide-Backing_Up.html#Administration_Guide-BU_Database (Which the RHEV proceadure BTW has a syntax error): psql -U postgres -d rhevm -W -f /usr/share/rhevm/db-backups/dump_RHEVDB_BACKUP_`date +%Y%m%d_%R`.sql should be: psql -U postgres -W -f /usr/share/rhevm/db-backups/dump_RHEVDB_BACKUP_`date +%Y%m%d_%R`.sql Either of the above procedures worked with nightly builds up until about a month back. I was able to backup a 3.0 database and restore it per the above on a system running the nightly build, then run the upgrade.sh script located in /usr/share/ovirt-engine/dbscripts. There is no script in /usr/share/ovirt-engine/dbscripts called upgrade_db.sh I always ran upgrade.sh -u postgres and it worked fine. I am not using create_db.sh script in /usr/share/ovirt-engine/dbscirpts to create the database and objects. The createdb command I use is the standard postgres command to create an new empty database. (/bin/createdb). Thus what imports is actually a clean restore of the 3.0 database Similarly I am using /bin/dropdb to destroy/delete the engine database. Thus importing the dump from 3.0 works using those procedures but results in the previous error I noted when attempting to run /usr/share/ovirt-engine/dbscripts/upgrade.sh. - DHC On Thu, Jul 19, 2012 at 7:04 AM, Robert Middleswarth rob...@middleswarth.net wrote: On 07/19/2012 04:10 AM, Eli Mesika wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com** To: users@ovirt.org Sent: Thursday, July 19, 2012 2:22:10 AM Subject: Re: [Users] ovirt-engine upgrade 3.0 to 3.1 issue Forgot reply-all sending to ovirt-users as well. -DHC On Wed, Jul 18, 2012 at 6:03 PM, Dead Horse deadhorseconsult...@gmail.com wrote: 3.0 engine database dump attached as: engine.sql.tar.gz - DHC On Wed, Jul 18, 2012 at 1:05 PM, Eli Mesika emes...@redhat.com wrote: - Original Message - From: Dead Horse deadhorseconsult...@gmail.com To: Users@ovirt.org Sent: Wednesday, July 18, 2012 9:08:55 AM Subject: [Users] ovirt-engine upgrade 3.0 to 3.1 issue Steps taken: Load up bare metal or a VM with FC16 Install 3.0 from http://www.ovirt.org/releases/**stable/fedora/16/http://www.ovirt.org/releases/stable/fedora/16/ Setup up something minimal (EG: engine-setup then setup a basic datacenter/cluster/etc) Add a FC16 or EL based node for fun as well and some VM's if feeling ambitious. Back up database: systemctl stop jboss-as.service pg_dump -C -E UTF8 --column-inserts --disable-dollar-quoting --disable-triggers -U postgres --format=p -f /temp/engine.sql engine Back up the following files within /etc/pki/ovirt-engine from a 3.0 install (preserve owner/group). /etc/pki/ovirt-engine/cacert.**conf /etc/pki/ovirt-engine/ca.pem /etc/pki/ovirt-engine/cert.**conf /etc/pki/ovirt-engine/certs/**01.pem /etc/pki/ovirt-engine/certs/**02.pem /etc/pki/ovirt-engine/certs/**ca.der /etc/pki/ovirt-engine/certs/**engine.cer /etc/pki/ovirt-engine/certs/**engine.der /etc/pki/ovirt-engine/**database.txt /etc/pki/ovirt-engine/**database.txt.attr /etc/pki/ovirt-engine/**database.txt.attr.old /etc/pki/ovirt-engine/**database.txt.old /etc/pki/ovirt-engine/keys/**engine_id_rsa /etc/pki/ovirt-engine/keys/**engine.ssh.key.txt /etc/pki/ovirt-engine/private/**ca.pem /etc/pki/ovirt-engine/**requests/ca.csr /etc/pki/ovirt-engine/**requests/engine.req Load up bare metal or a VM with FC17 (Upgrade from FC16 -- FC17 is still rather messy) Install 3.1 from http://www.ovirt.org/releases/**beta/fedora/17/http://www.ovirt.org/releases/beta/fedora/17/ Run engine-setup to get an initial setup. stop ovirt-engine (systemctl stop ovirt-engine.service) Copy the previously backed up /etc/pki/ovirt-engine files from above over top of the ones just generated (be sure to preserve owner/group EG: cp -a) Drop the existing engine database: dropdb -U postgres engine Create a new blank database: createdb -U postgres engine Well, the problem is that you are creating here the database with createdb. This creates all DB objects and runs all upgrade scripts. Then when you try to restore from your file , you have problems, since your file also tries to create the same objects. instead of : createdb -U postgres engine do : echo create database engine; | psql -U postgres template1 Then continue on the next steps and it will work. optionally, in your backup command: instead of pg_dump -C -E UTF8 --column-inserts --disable-dollar-quoting use pg_dump --data-only -E UTF8 --column-inserts --disable-dollar-quoting then , createdb will create the db schema and your engine.sql will include only insert Eli I am anything but a PG expert but exporting only the data
[Users] ovirt-engine upgrade 3.0 to 3.1 issue
Steps taken: Load up bare metal or a VM with FC16 Install 3.0 from http://www.ovirt.org/releases/stable/fedora/16/ Setup up something minimal (EG: engine-setup then setup a basic datacenter/cluster/etc) Add a FC16 or EL based node for fun as well and some VM's if feeling ambitious. Back up database: systemctl stop jboss-as.service pg_dump -C -E UTF8 --column-inserts --disable-dollar-quoting --disable-triggers -U postgres --format=p -f /temp/engine.sql engine Back up the following files within /etc/pki/ovirt-engine from a 3.0 install (preserve owner/group). /etc/pki/ovirt-engine/cacert.conf /etc/pki/ovirt-engine/ca.pem /etc/pki/ovirt-engine/cert.conf /etc/pki/ovirt-engine/certs/01.pem /etc/pki/ovirt-engine/certs/02.pem /etc/pki/ovirt-engine/certs/ca.der /etc/pki/ovirt-engine/certs/engine.cer /etc/pki/ovirt-engine/certs/engine.der /etc/pki/ovirt-engine/database.txt /etc/pki/ovirt-engine/database.txt.attr /etc/pki/ovirt-engine/database.txt.attr.old /etc/pki/ovirt-engine/database.txt.old /etc/pki/ovirt-engine/keys/engine_id_rsa /etc/pki/ovirt-engine/keys/engine.ssh.key.txt /etc/pki/ovirt-engine/private/ca.pem /etc/pki/ovirt-engine/requests/ca.csr /etc/pki/ovirt-engine/requests/engine.req Load up bare metal or a VM with FC17 (Upgrade from FC16 -- FC17 is still rather messy) Install 3.1 from http://www.ovirt.org/releases/beta/fedora/17/ Run engine-setup to get an initial setup. stop ovirt-engine (systemctl stop ovirt-engine.service) Copy the previously backed up /etc/pki/ovirt-engine files from above over top of the ones just generated (be sure to preserve owner/group EG: cp -a) Drop the existing engine database: dropdb -U postgres engine Create a new blank database: createdb -U postgres engine Populate with with the backup of the above 3.0 database: psql -U postgres -d engine -w /temp/engine.sql Attempt to upgrade the DBschema via /usr/share/ovirt-engine/dbscripts/upgrade.sh upgrade.sh -u postgres The DB schema upgrade will fail like so: Running upgrade script upgrade/03_01_0260_add_job_table.sql psql:upgrade/03_01_0260_add_job_table.sql:29: NOTICE: CREATE TABLE / PRIMARY KEY will create implicit index pk_jobs for table job CONTEXT: SQL statement CREATE TABLE job ( job_id UUID NOT NULL, action_type VARCHAR(50) NOT NULL, description TEXT NOT NULL, status VARCHAR(32) NOT NULL, owner_id UUID, visible BOOLEAN NOT NULL DEFAULT true, start_time TIMESTAMP WITH TIME ZONE NOT NULL, end_time TIMESTAMP WITH TIME ZONE default NULL, last_update_time TIMESTAMP WITH TIME ZONE default NULL, correlation_id VARCHAR(50) NOT NULL, CONSTRAINT pk_jobs PRIMARY KEY(job_id) ) WITH OIDS PL/pgSQL function __temp_upgrade_add_job_table line 6 at SQL statement psql:/tmp/tmp.mXz8U4xpWr:3: ERROR: column comment of relation schema_version does not exist LINE 1: ...ksum,installed_by,started_at,ended_at,state,current,comment) My assumption here is that if the DB schema upgrade would work, that this should be the cleanest way to upgrade from 3.0 to 3.1. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
Re: [Users] ovirt and NAT/Internal Networks
Did some further refining and testing on this. I found that all that needed to be done to create a NAT or private network for your ovirt guests is (in this example we create a NAT network): 1) Login into your ovirt nodes (all will need to have this created consistently across them. 2) connect to qemu/libvirt with virsh 3) virsh# connect qemu:///system 4) username is: vdsm@rhevh 5) password is found in the file /etc/pki/vdsm/keys/libvirt_password 6) Create a network xml definition like so (in this example NAT): network namenat/name uuidb09d09a8-ebbd-476d-9045-e66012c9e83d/uuid forward mode='nat'/ bridge name='natbr0' stp='on' delay='0' / mac address='52:54:00:9D:82:DE'/ ip address='192.168.1.1' netmask='255.255.255.0' dhcp range start='192.168.1.2' end='192.168.1.250' / /dhcp /ip /network 7) Save the network definition at /etc/libvirt/qemu/networks (EX: /etc/libvirt/qemu/networks/nat.xml) 8) Add the network: virsh# net-define /etc/libvirt/qemu/networks/nat.xml) 9) Set the network to autostart: virsh# net-autostart nat.xml 10) If this is NAT you will need to enable packet forwarding in /etc/sysctl.conf (EG: set net.ipv4.ip_forward = 1) 11) If you don't want the reboot the host for this to take effect do: echo '1' /proc/sys/net/ipv4/ip_forward 12) In the ovirt UI select the cluster which contains the hosts on which you created the network 13) Create a new logical network with the same name as the bridge name element from the above xml.(EX: natbr0) 14) Remember to enable STP in this case since this is a private or NAT network 15) The logical network will be show as non-operational, so we will need to manually online the network. -- vdsm/ovirt network creation seems to be tied to host physical interfaces and does not look at virtual interfaces in this case vdsm kicks in and finds natbr0 and consequently creates vdsm-natbr0 and natbr0-nic but does nothing more to report it back to the ovirt engine. (ovirt developers feel free to comment on this...) 16) Thus since ovirt does not detect that natbr0 exists and is actually online we need to make a quick database modification to force it online so it can actually be used. 17) We first need to find out the UUID of natbr0 interface 18) type the following: psql -U postgres engine -c SELECT network.id FROM public.network WHERE network.name = 'natbr0'; (remember to substitute the name you used for the network) 20) This returns you the UUID of the network 21) Next we need to online the network now that we know it's UUID 22) type the following: psql -U postgres engine -c update network_cluster set status='1' where network_id='the network uuid returned from the previous command goes here'; 23) You should now be able to attach guest vnics to your newly created NAT or private network. oVirt developers, any hope of getting this supported officially as part of ovirt? It seems like it would be relatively easy since the functionality exists in libvirt, and vdsm seems to be aware but the hooks into the engine/UI are what seem to be missing. - DHC On Mon, Apr 23, 2012 at 4:49 PM, Dead Horse deadhorseconsult...@gmail.comwrote: I have tried to create a NAT network within ovirt for guests to use via libvirt/virsh as the vdsm@rhevh on an ovirt node. Something like this: virsh # net-info nat Namenat UUIDb09d09a8-ebbd-476d-9045-e66012c9e83d Active: yes Persistent: yes Autostart: yes Bridge: natbr0 virsh # net-list --all Name State Autostart - nat active yes vdsm-natbr0 active yes vdsm-ovirtmgmt active yes virsh # virsh # net-edit nat network namenat/name uuidb09d09a8-ebbd-476d-9045-e66012c9e83d/uuid forward mode='nat'/ bridge name='natbr0' stp='on' delay='0' / mac address='52:54:00:9D:82:DE'/ ip address='192.168.122.1' netmask='255.255.255.0' dhcp range start='192.168.122.2' end='192.168.122.254' / /dhcp /ip /network Set net.ipv4.ip_forward = 1 in sysctl.conf This is basically the default libvirt NAT network. Connecting a VM directly to it via cli (qemu-kvm) works perfectly fine on the ovirt host itself but the ovirt manager UI does not seem to pick up on it or allow for creation of such a network for guests to use. Since the ovirt manager UI did not allow I dug into the database and added entries which seem to have been necessary by hand to the following engine database tables: table: network table: network_cluster table: vds_interface table: vds_interface_statistics The result was the network did show up under cluster logical networks. However the network still does not show up under the host Network Interfaces UI. Was able to create VM and attach the natbr0 network to it and NAT was indeed working. (EG: DNSmasq assigned address). The guest was able to communicate out and without
[Users] ovirt and NAT/Internal Networks
I have tried to create a NAT network within ovirt for guests to use via libvirt/virsh as the vdsm@rhevh on an ovirt node. Something like this: virsh # net-info nat Namenat UUIDb09d09a8-ebbd-476d-9045-e66012c9e83d Active: yes Persistent: yes Autostart: yes Bridge: natbr0 virsh # net-list --all Name State Autostart - nat active yes vdsm-natbr0 active yes vdsm-ovirtmgmt active yes virsh # virsh # net-edit nat network namenat/name uuidb09d09a8-ebbd-476d-9045-e66012c9e83d/uuid forward mode='nat'/ bridge name='natbr0' stp='on' delay='0' / mac address='52:54:00:9D:82:DE'/ ip address='192.168.122.1' netmask='255.255.255.0' dhcp range start='192.168.122.2' end='192.168.122.254' / /dhcp /ip /network Set net.ipv4.ip_forward = 1 in sysctl.conf This is basically the default libvirt NAT network. Connecting a VM directly to it via cli (qemu-kvm) works perfectly fine on the ovirt host itself but the ovirt manager UI does not seem to pick up on it or allow for creation of such a network for guests to use. Since the ovirt manager UI did not allow I dug into the database and added entries which seem to have been necessary by hand to the following engine database tables: table: network table: network_cluster table: vds_interface table: vds_interface_statistics The result was the network did show up under cluster logical networks. However the network still does not show up under the host Network Interfaces UI. Was able to create VM and attach the natbr0 network to it and NAT was indeed working. (EG: DNSmasq assigned address). The guest was able to communicate out and without extensive testing things seemed to work. This was very painful to do (and hacky). Thus it would be very nice if since this is OOB functionality of libvirt/KVM, if ovirt would allow for it to be used. The ability to provide NAT or internal networks to guests would quite useful. - DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
[Users] fedora qemu modern CPU recognition issues cause ovirt manged host to go non-operational
Seeing an issue wherein ovirt moves a managed host to non-operational state. This occurs with the currently released version of ovirt and the latest development builds. The ovirt host is loaded with Fedora core 16 and equipped with most current development version of the vdsm. *Editorial node* The latest vdsm to work on the FC16 host required building and adding newer versions of the sanlock, libvirt, lvm2 and device-mapper packages than what FC16 provides. Ultimately however none of the newer packages have any bearing on this failure mode. The failure mode is as follows. Upon successfully adding the host and setting the cluster CPU compatibility level oVirt will offline the host with the following message: -- Host ovirtnode moved to Non-Operational state as host does not meet the cluster's minimum CPU level. Missing CPU features : model_Nehalem Under the hood the actual cause of this failure is that qemu is not correctly able to identify the host CPU feature flags. This can be observed by doing: qemu-system-x86_64 -cpu Nehalem,check which fails with: warning: host cpuid _ lacks requested flag 'fpu' [0x0001] warning: host cpuid _ lacks requested flag 'de' [0x0004] and on and on... A simple check of cat /proc/cpuinfo | grep flags flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 popcnt lahf_lm ida dts tpr_shadow vnmi flexpriority ept vpid Thus the CPU is more than capable of everything being asked of it via the cpudef for Nehalem in /etc/qemu/target-x86_64.conf. This is only an issue on Fedora hosts. RHEL/CentOS/SL hosts work fine. This was recognized as an issue RHEL and fixed there but has not been fixed in Fedora. See: http://wiki.qemu.org/Features/CPUModels#Examples and this: https://bugzilla.redhat.com/show_bug.cgi?format=multipleid=689665 and this related discussion in qemu-devel: http://www.mail-archive.com/qemu-devel@nongnu.org/msg101360.html Thus is appears that the changes were made to the RHEL qemu (eg: cpu type rhel6) AKA the change one needs to make to the ovirt engine database to have ovirt manage a EL based host. Fedora hosts-- psql -U postgres engine -c update vdc_options set option_value='pc-0.14' where option_name='EmulatedMachine' and version='3.0'; EL hosts -- psql -U postgres engine -c update vdc_options set option_value='rhel6.2.0' where option_name='EmulatedMachine' and version='3.0'; Thus at the moment any host loaded with Fedora and manged by oVirt utilizing a Sandy Bridge, Nehalem or Westmere processor will be dead in the water. -DHC ___ Users mailing list Users@ovirt.org http://lists.ovirt.org/mailman/listinfo/users
