Re: [one-users] Unable to auto start one and sunstone service in Ubuntu 13.10

[Stefan Kooman]

Quoting kiran ranjane (kiran.ranj...@gmail.com):
 
> I have tested this and it works well. I get only 3 to 4 timeout request
> when fail-over is triggered so this is quite instant and simple to
> troubleshoot in-case of issues, No split brain as M/M replication is used
> and both the database and always in sync, Mostly all the sync is from one
> side Server A to B because server A is the master node and services are
> running on one server at a time. Services on Server B starts only after
> fail-over is triggered so the syncing of data from B to A is always less
> because it is used as standby cloud storage server.

You state "No split brain as M/M replication is used ...". But what if
communication between server A and B is lost? Server A becomes master
and starts one. Server B becomes master and starts one. They both update
the one database, but replication is not (yet) happening because no
communication is possible so no issue yet. After communication is
restored replication continues, and this might work without errors
(although I have my doubts about that). One master node (B) will proably
shuts itself down. Master A continues to run. I wonder how healthy the
one database is after such an incident. Do you do manual fail-overs or
automatic ones?

Gr. Stefan


-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] one-context RPM not working on Fedora 20

[ML mail]

Hi Stefan,

Ah that makes sense, thanks for clarifying! Have a productive and fun day today 
at the dojo ;)

Cheers

M.L.




On Thursday, January 30, 2014 9:43 PM, Stefan Kooman  wrote:
 
Quoting ML mail (mlnos...@yahoo.com):
> Hello,
> 
> I would like to contextualize a Fedora 20 desktop VM. For that purpose
> I used the offical one-context package from OpenNebula
> (http://dev.opennebula.org/attachments/download/747/one-context_4.4.0.rpm)
> but it simply does not work, my eth0 interface does not get any IP
> address automatically so that eth0 stays without an IP.
> 
> Is this package maybe known not to work on Fedora? or am I missing
> something?

Fedora 20 is using systemd [1], which is not yet supported by
one-context.  But ... at the CentOS dojo tomorrow in Brussels the CentOS
and Opennebula team try to hack together systemd compatible context
scripts [2]. You just have to wait one more day :).


Gr. Stefan

[1]: https://fedoraproject.org/wiki/Systemd

[2]: http://wiki.centos.org/Events/Dojo/Brussels2014


-- 
| BIT BV  http://www.bit.nl/        Kamer van Koophandel 09090351
| GPG: 0xD14839C6                   +31 318 648 688 / i...@bit.nl___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

[one-users] vm memory usage

[kenny . kenny]

 
Is possible to get the guest memory usage by one command line or by xen ?
 
thanks.
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Unable to auto start one and sunstone service in Ubuntu 13.10

[kiran ranjane]

Hi Jaime,

The setup is quite simple, Tools needed are ucarp, rsync, Mysql
master-master replication.

We need to use a virtual IP which is configured using ucarp and that
virtual IP can be used to access Sunstone. I have setup this with 2 servers
but can configure ucarp with 3 or 4 servers. We need to modify the VIP-up
and VIP-down script in ucarp by adding "One" and Sunstone start script in
vip-up and one and Sun-stone stop scrip in vip-down script of ucarp. These
script makes sure that the IP is fail-over and fail-back between nodes and
also the one services are started or stopped on the nodes accordingly while
fail-over or fail-back happens.

Rsync is used to sync "ONE" (self-contained mode) folder from SERVER
A(MASTER NODE) to SERVER B , You can setup a cron job to sync the folder
once a day.

You can also use system-wide mode but you need to make sure all the folder
are synced between both the servers which are located in different location
in system-wide mode and also you will not face the start-up issue that I am
facing in contained-mode. However keep in mind you also need to exclude few
folders in rsync command so that they should not be synced between servers
like the var/datastore and logs files.
Open-nebula and mysql should be install on both the servers and they need
to be in sync from Server A to Server B. Open-nebula can be synced using
rync command and Mysql through Master-Master replication.

Next step is to create password-less ssh between serverA and serverB  and
password-less ssh between serverA and all kvm nodes, serverB and all KVM
node. (In my setup I have dedicated KVM nodes and dedicated cloud storage
management nodes which have Open-nebula and glustefs installed on same
node) So that when fail-over is triggered the ONE Server is able to
communicate from Server B with all the KVM nodes.

After you have installed open-nebula and configured rsync properly then we
can start with MYSQL Master-Master replication from Server A to Server B.
You need to replicated the database that was created while open-nebula was
installed. Make sure you check mysql version before setting up the M/M
replication because there are few variation in setup and parameters in
config file after 5.5 and up.

After the M/M replication is setup we need to make sure that it is always
running from server A to Server B and from server B to server A. We have
setup M/M replication so that any changes on open-nebula DB on server A is
synced on server B.

Next we need to make sure that "One" and Sunstone service are not running
on Server B at all that means it should also not run in start-up. Only
MySQL service should be running on server B, When fail over is triggered
ucarp vip-up script assigns virtual Ip to server B and then starts "one"
and Sunstone service on Server B and when fail-back is triggered then
vip-down scrip will stop the services and removes the virtual ip and
assigns it back to master node (Server A).

I have tested this and it works well. I get only 3 to 4 timeout request
when fail-over is triggered so this is quite instant and simple to
troubleshoot in-case of issues, No split brain as M/M replication is used
and both the database and always in sync, Mostly all the sync is from one
side Server A to B because server A is the master node and services are
running on one server at a time. Services on Server B starts only after
fail-over is triggered so the syncing of data from B to A is always less
because it is used as standby cloud storage server.

This Setup is simple and works well and do not need any complex packages
like lucci/ricci or corosync/pacemaker. Ucarp is also available on Centos
and Fedora you can also try on them

If you have any questions then you can get back to me.

Thanks and Regards
Kiran Ranjane



On Thu, Jan 30, 2014 at 2:59 PM, Jaime Melis  wrote:

> Hi Kiran,
>
> That sounds very interesting, at some point it would be great if you could
> elaborate a bit more on exactly how you have done it, but it looks really
> nice.
>
> The thing about self contained mode is that it's desgined for development
> or for custom solutions, such as yours. It's not intented for packaging or
> to provide commercial supoprt.
>
> So, in your particular case, I think you can fix this quite easily.
> Continue with the self-contained mode, but instead of copying those scripts
> (one and sunstone-server), install these files to /etc/init.d:
> https://github.com/OpenNebula/one/blob/master/share/pkgs/Ubuntu/opennebula
>
> https://github.com/OpenNebula/one/blob/master/share/pkgs/Ubuntu/opennebula-sunstone
>
> Then, make some changes to them so they work for system contained mode. It
> shouldn't be much change, maybe just sourcing the file were you have the
> ONE_LOCATION variable defined before doing one start and one stop.
>
> Let us know how it goes.
>
> Cheers,
> Jaime
>
>
>
>
>
>
> On Mon, Jan 27, 2014 at 8:06 PM, kiran ranjane wrote:
>
>> Hi Jaime,
>>
>> May I know which version of Ubuntu is officially s

Re: [one-users] one-context RPM not working on Fedora 20

[Stefan Kooman]

Quoting ML mail (mlnos...@yahoo.com):
> Hello,
> 
> I would like to contextualize a Fedora 20 desktop VM. For that purpose
> I used the offical one-context package from OpenNebula
> (http://dev.opennebula.org/attachments/download/747/one-context_4.4.0.rpm)
> but it simply does not work, my eth0 interface does not get any IP
> address automatically so that eth0 stays without an IP.
> 
> Is this package maybe known not to work on Fedora? or am I missing
> something?

Fedora 20 is using systemd [1], which is not yet supported by
one-context.  But ... at the CentOS dojo tomorrow in Brussels the CentOS
and Opennebula team try to hack together systemd compatible context
scripts [2]. You just have to wait one more day :).


Gr. Stefan

[1]: https://fedoraproject.org/wiki/Systemd
[2]: http://wiki.centos.org/Events/Dojo/Brussels2014


-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

[one-users] one-context RPM not working on Fedora 20

[ML mail]

Hello,

I would like to contextualize a Fedora 20 desktop VM. For that purpose I used 
the offical one-context package from OpenNebula 
(http://dev.opennebula.org/attachments/download/747/one-context_4.4.0.rpm) but 
it simply does not work, my eth0 interface does not get any IP address 
automatically so that eth0 stays without an IP.

Is this package maybe known not to work on Fedora? or am I missing something?

Regards,
M.L.___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] oneacct [VirtualMachinePoolAccounting] Internal Error

[Carlos Martín Sánchez]

Hi Stefan,

On Tue, Jan 28, 2014 at 5:57 PM, Stefan Kooman  wrote:

> Hi list,
>
> Running oneact as oneadmin user I get the following error:
>
> oneacct
> [VirtualMachinePoolAccounting] Internal Error.
>
> oned.log shows:
>
> Tue Jan 28 17:21:19 2014 [ReM][D]: Req:6624 UID:0
> VirtualMachinePoolAccounting invoked, -2, -1, -1
> Tue Jan 28 17:21:19 2014 [ONE][E]: SQL command was: SELECT history.body
> FROM history INNER JOIN vm_pool WHERE vid=oid GROUP BY vid,seq, error
> 1055 : 'opennebula.history.body' isn't in GROUP BY
> Tue Jan 28 17:21:19 2014 [ReM][E]: Req:6624 UID:0
> VirtualMachinePoolAccounting result FAILURE
> [VirtualMachinePoolAccounting] Internal Error
>
> I don't know for sure it ever worked after I upgraded to onennebula-4.4.
>
> It might have to do with our mysql server settings (sql-mode):
> sql-mode   =
>
> STRICT_TRANS_TABLES,ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER,NO_AUTO_VALUE_ON_ZERO,NO_ENGINE_SUBSTITUTION,NO_ZERO_DATE,NO_ZERO_IN_DATE,ONLY_FULL_GROUP_BY
>
> But I'm not sure.
>
> How to best debug this issue?
>
> Gr. Stefan
>

It definitely looks like a problem with the ONLY_FULL_GROUP_BY flag.
Now that I think about it, we don't really need the group by, and it may
probably be a bug and we meant to use ORDER BY [1].

Regards

[1] http://dev.opennebula.org/issues/2700

--
Carlos Martín, MSc
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org  | cmar...@opennebula.org |
@OpenNebula 
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] VNC in sunstone not working on firefox 26 (one 4.4)

[Hamada, Ondrej]

Hi Tino,
Thank you for reply. Here's the output:

17:45:06.596 POST https://opennebulaaddr/vm/272/startvnc [HTTP/1.1 200 OK 59ms]
17:45:06.621 SecurityError: The operation is insecure. websock.js:333
17:45:06.619 "New state 'loaded', was 'disconnected'. Msg: noVNC ready: native 
WebSockets, canvas rendering" util.js:110
17:45:06.620 "New state 'connect', was 'loaded'." util.js:110
17:45:06.621 "Skipping unsupported WebSocket binary sub-protocol" util.js:111
17:45:08.621 "New state 'failed', was 'connect'. Msg: Connect timeout" 
util.js:111
17:45:08.672 "New state 'disconnected', was 'failed'."

The security error - it remains me that the problems had started probably after 
I've configured apache as a SSL proxy for sunstone. And in Chrome it works ok. 
Unfortunately I need the SSL enabled.

-Original Message-
From: Tino Vazquez [mailto:cvazq...@c12g.com]
Sent: Thursday, January 30, 2014 4:24 PM
To: Hamada, Ondrej
Cc: users
Subject: Re: [one-users] VNC in sunstone not working on firefox 26 (one 4.4)

Hi Ondrej,

Right after clicking on the VNC link, is anything showing in the Firefox dev 
tools console [1]?

Best,

-Tino

[1] https://developer.mozilla.org/en/docs/Tools
--
OpenNebula - Flexible Enterprise Cloud Made Simple

--
Constantino Vázquez Blanco, PhD, MSc
Senior Infrastructure Architect at C12G Labs www.c12g.com | @C12G | 
es.linkedin.com/in/tinova

--
Confidentiality Warning: The information contained in this e-mail and any 
accompanying documents, unless otherwise expressly indicated, is confidential 
and privileged, and is intended solely for the person and/or entity to whom it 
is addressed (i.e. those identified in the "To" and "cc" box). They are the 
property of C12G Labs S.L..
Unauthorized distribution, review, use, disclosure, or copying of this 
communication, or any part thereof, is strictly prohibited and may be unlawful. 
If you have received this e-mail in error, please notify us immediately by 
e-mail at ab...@c12g.com and delete the e-mail and attachments and any copy 
from your system. C12G thanks you for your cooperation.


On 27 January 2014 18:27, Ondrej Hamada  wrote:
> Hi,
> currently I'm unable to use VNC from sunstone (OpenNebula 4.4, debian
> 7.2). Sunstone shows the information about VNC start and then no VNC
> windows pops up. This is happening on firefox 26 only (and also on my
> colleague's computer). The same web in chromium 32 works ok. Has
> anyone else encountered same problem? Any idea where's the problem?
>
> --
> Ondrej Hamada
>
>
> This e-mail and any attachment is for authorised use by the intended
> recipient(s) only. It may contain proprietary material, confidential
> information and/or be subject to legal privilege. It should not be
> copied, disclosed to, retained or used by, any other party. If you are
> not an intended recipient then please promptly delete this e-mail and
> any attachment and all copies and inform the sender. Thank you for
> understanding.
>
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org


This e-mail and any attachment is for authorised use by the intended 
recipient(s) only. It may contain proprietary material, confidential 
information and/or be subject to legal privilege. It should not be copied, 
disclosed to, retained or used by, any other party. If you are not an intended 
recipient then please promptly delete this e-mail and any attachment and all 
copies and inform the sender. Thank you for understanding.


___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] onevnet policy violation?

[Carlos Martín Sánchez]

Hi Stefan,

On Thu, Jan 30, 2014 at 9:18 AM, Stefan Kooman  wrote:

> Hi,
>
> I created a new vnet based on the following template:
>
> --- template start ---
> NAME= "v-test01"
> TYPE= "ranged"
> BRIDGE  = "uplink"
> CLUSTER = "kvm_cluster"
>
> NETWORK_ADDRESS = 172.17.226.0
> NETWORK_MASK= 255.255.255.224
> IP_START= 172.17.226.4
> IP_END  = 172.17.226.30
>
> GATEWAY = 172.17.226.1
>
> VLAN= "YES"
> VLAN_ID = 226
> --- template end ---
>
> After creation the vnet is not a member of the kvm_cluster but instead
> of the default cluster "-". Is this a bug or is it me doing it wrong?
> After reading [1] again it explicitly states a "vnet or datastore"
> should be added with the onecluster add command, so I was doing it
> wrong. It still would be nice to have it automatically added to a
> cluster if "CLUSTER" attribute is there.
>

You can create vnets directly in a cluster, using the --cluster option [1],
but not with the CLUSTER attribute [2].
I agree that since the new vnet is defined with a template, it is more
intuitive to set the cluster also there, so I've opened a ticket for this
[3].


> After that  I instantiated a vm based on a template with a mix of
> resources from "kvm_cluster" and "-" (v-test01). This vm succesfully
> managed to get into running state. According to [1] section "The Default
> Cluster 'None'" hosts in the Cluster 'none' will only run VMs using
> resources without a cluster. I assume this is equally true for hosts
> that are a member of a non-default cluster (kvm_cluster in our case) but
> instead it did run the vm. Indeed, according to "The "Scheduling and
> Cluster" section of [1] the deployment should have failed. What info do
> we need to make sure it's a bug and not PEBKAC?
>
> Gr. Stefan
>
> [1]:
>
> http://docs.opennebula.org/4.4/administration/hosts_and_clusters/cluster_guide.html


This is a bit hard to explain easily in the docs, but the key is this
paragraph [4]:

Virtual Machines using resources from Datastores or Virtual Networks in the
Cluster 'none' can be deployed in any Host, which must be properly
configured.

That is, your vnet is not in any cluster (or cluster 'none'), which means
that all hosts (either in cluster none or any other cluster) will be able
to deploy the VM.
The core sees that you are requesting something specific to cluster
kvm_cluster and a vnet that is configured to be used by any Host, including
those in kvm_cluster.

If you try to create a VM using resources from two different clusters, the
core will complain on creation, it won't wait for the scheduler to notice.


This may be a bit convoluted, but the other option we had was to force you
to put everything in a cluster. That would mean that you can't configure a
network to be used by both xen and kvm hosts.
In future versions, we may change this and allow to select more than one
cluster for each host/vnet/ds...

Regards

[1] http://docs.opennebula.org/doc/4.4/cli/onevnet.1.html
[2] http://docs.opennebula.org/4.4/user/references/vnet_template.html
[3] http://dev.opennebula.org/issues/2698
[4]
http://docs.opennebula.org/4.4/administration/hosts_and_clusters/cluster_guide.html#the-default-cluster-none
--
Carlos Martín, MSc
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org  | cmar...@opennebula.org |
@OpenNebula 
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] EC2 / cloud bursting - multiple AWS credentials

[Carlos Martín Sánchez]

Hi Stefan,

On Thu, Jan 30, 2014 at 7:52 AM, Stefan Kooman  wrote:

> Hi,
>
> I was reading through Amazon EC2 prerequisites [1] which implies that
> there can be only one set of AWS credentials per opennebula cloud. Is
> that correct? This might not be a problem for a "private cloud" operated
> by only one organisation / company. For a public cloud that want's to
> leave room for 3rd party cloud bursting it is a problem. Ideally every
> user / group should be able to provide his/her own credentials while
> instantiating/creating a new vm. What is the reason to use a config file
> for this instead of having this info in a template?
>
> Gr. Stefan
>
> [1]:
>
> http://docs.opennebula.org/4.4/advanced_administration/cloud_bursting/ec2g.html#prerequisites


Actually you can define multiple ec2 accounts, see the Multi EC2
Site/Region/Account section of that guide [1]. You can create a hybrid host
for each group, and then adjust the permissions so each one can only deploy
VMs in the host with the right credentials.

We could come up with an ec2 driver that reads the credentials from the VM
template, although I'm not sure how difficult it would be to make it work
with the current code. Please open a feature request if the above multi
account feature does not solve your use case.

Regards

[1]
http://docs.opennebula.org/4.4/advanced_administration/cloud_bursting/ec2g.html#multi-ec2-site-region-account-support

--
Carlos Martín, MSc
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org  | cmar...@opennebula.org |
@OpenNebula 
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] VNC in sunstone not working on firefox 26 (one 4.4)

[Tino Vazquez]

Hi Ondrej,

Right after clicking on the VNC link, is anything showing in the
Firefox dev tools console [1]?

Best,

-Tino

[1] https://developer.mozilla.org/en/docs/Tools
--
OpenNebula - Flexible Enterprise Cloud Made Simple

--
Constantino Vázquez Blanco, PhD, MSc
Senior Infrastructure Architect at C12G Labs
www.c12g.com | @C12G | es.linkedin.com/in/tinova

--
Confidentiality Warning: The information contained in this e-mail and
any accompanying documents, unless otherwise expressly indicated, is
confidential and privileged, and is intended solely for the person
and/or entity to whom it is addressed (i.e. those identified in the
"To" and "cc" box). They are the property of C12G Labs S.L..
Unauthorized distribution, review, use, disclosure, or copying of this
communication, or any part thereof, is strictly prohibited and may be
unlawful. If you have received this e-mail in error, please notify us
immediately by e-mail at ab...@c12g.com and delete the e-mail and
attachments and any copy from your system. C12G thanks you for your
cooperation.


On 27 January 2014 18:27, Ondrej Hamada  wrote:
> Hi,
> currently I'm unable to use VNC from sunstone (OpenNebula 4.4, debian
> 7.2). Sunstone shows the information about VNC start and then no VNC
> windows pops up. This is happening on firefox 26 only (and also on my
> colleague's computer). The same web in chromium 32 works ok. Has anyone
> else encountered same problem? Any idea where's the problem?
>
> --
> Ondrej Hamada
>
>
> This e-mail and any attachment is for authorised use by the intended
> recipient(s) only. It may contain proprietary material, confidential
> information and/or be subject to legal privilege. It should not be copied,
> disclosed to, retained or used by, any other party. If you are not an
> intended recipient then please promptly delete this e-mail and any
> attachment and all copies and inform the sender. Thank you for
> understanding.
>
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Gareth Bult]

Sorry, my fault, attention is elsewhere ..

I actually use "writethrough" rather than "writeback" .. so with the exception 
of the quote from the
mailing list, where I said "writethrough", please read "writeback" ... 
apologies.

I "used" to use writeback, but then read a few horror stories and tried 
writethrough .. and when used
with vdc-store, writethrough .vs. writeback makes almost no difference anyway 
as vdc-store buffers the
output "and" is migration safe ...

I've done (literally) thousands of migrations using VM's under IO load with 
writethrough cache (damn! 
I just typed  writeback again and had to correct! losing my mind!) while 
testing with ON, never had a 
problem or corruption - so for me writethrough is certainly safe enough to make 
me want to use the
sharable flag .. :)

-- 

Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 
See the status of my current project at http://vdc-store.com 


- Original Message -
From: "Stefan Kooman" 
To: "Gareth Bult" 
Cc: "Javier Fontan" , users@lists.opennebula.org
Sent: Thursday, 30 January, 2014 2:59:33 PM
Subject: Re: [one-users] Setting the "SHAREABLE" attribute

Quoting Gareth Bult (gar...@linux.co.uk):
> Mmm, there has been some discussion around this ..
> 
> Technically using cache=writeback "should" be safe for migration.
> (which is the method I use)
> 
> So .. I need "cache=writeback" and "sharable" for migration to happen.
> (the performance hit with cache=off is an unacceptable performance hit, at 
> least for me ..)
> 
> From the QEMU lists, 2012;
> "In short, if you're using a recent kernel with ext3 or ext4, cache=writeback 
> is absolutely safe. If you're using an older version of ext3, cache=writeback 
> is still safe but ext3 itself isn't. cache=writeback can make the situation 
> worse."

Hmm, the guys at big blue think otherwise [1]:

With caching set to writeback mode, both the host page cache and the
disk write cache are enabled for the guest. Because of this, the I/O
performance for applications running in the guest is good, but the data
is not protected in a power failure. As a result, this caching mode is
recommended only for temporary data where potential data loss is not a
concern.

Gr. Stefan

[1]:
http://pic.dhe.ibm.com/infocenter/lnxinfo/v3r0m0/index.jsp?topic=%2Fliaat%2Fliaatbpkvmasynchio.htm

-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Stefan Kooman]

Quoting Gareth Bult (gar...@linux.co.uk):
> Mmm, there has been some discussion around this ..
> 
> Technically using cache=writeback "should" be safe for migration.
> (which is the method I use)
> 
> So .. I need "cache=writeback" and "sharable" for migration to happen.
> (the performance hit with cache=off is an unacceptable performance hit, at 
> least for me ..)
> 
> From the QEMU lists, 2012;
> "In short, if you're using a recent kernel with ext3 or ext4, cache=writeback 
> is absolutely safe. If you're using an older version of ext3, cache=writeback 
> is still safe but ext3 itself isn't. cache=writeback can make the situation 
> worse."

Hmm, the guys at big blue think otherwise [1]:

With caching set to writeback mode, both the host page cache and the
disk write cache are enabled for the guest. Because of this, the I/O
performance for applications running in the guest is good, but the data
is not protected in a power failure. As a result, this caching mode is
recommended only for temporary data where potential data loss is not a
concern.

Gr. Stefan

[1]:
http://pic.dhe.ibm.com/infocenter/lnxinfo/v3r0m0/index.jsp?topic=%2Fliaat%2Fliaatbpkvmasynchio.htm

-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Javier Fontan]

Reading that piece of code it seems I was totally wrong with that:

--8<--
if (disk->cachemode) {
// bunch of code
} else if (disk->shared && !disk->readonly) {
virBufferAddLit(&opt, ",cache=off");
}
-->8--

It should make some difference. We will take a look on the best way to
add the shareable flag and maybe others.

Cheers

On Thu, Jan 30, 2014 at 4:14 PM, Gareth Bult  wrote:
> Mmm, I can't disagree if that's what the code says .. however turning
> off the cache completely in the GUI does "seem" to make quite a *dramatic*
> difference to the VM performance, whereas setting the sharable flag seems
> to have no performance impact at all .. (when using writethrough ...)  ???
>
> --
>
> Gareth Bult
> "The odds of hitting your target go up dramatically when you aim at it."
> See the status of my current project at http://vdc-store.com
>
>
> - Original Message -
> From: "Javier Fontan" 
> To: "Gareth Bult" 
> Cc: "Stefan Kooman" , users@lists.opennebula.org
> Sent: Thursday, 30 January, 2014 2:48:15 PM
> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>
> I'll investigate a bit more on this issue but from the qemu driver
> code it seems that cache is disabled with shareable flag.
>
> On Thu, Jan 30, 2014 at 4:07 PM, Gareth Bult  wrote:
>> Mmm, there has been some discussion around this ..
>>
>> Technically using cache=writeback "should" be safe for migration.
>> (which is the method I use)
>>
>> So .. I need "cache=writeback" and "sharable" for migration to happen.
>> (the performance hit with cache=off is an unacceptable performance hit, at 
>> least for me ..)
>>
>> From the QEMU lists, 2012;
>> "In short, if you're using a recent kernel with ext3 or ext4, 
>> cache=writeback is absolutely safe. If you're using an older version of 
>> ext3, cache=writeback is still safe but ext3 itself isn't. cache=writeback 
>> can make the situation worse."
>>
>> ...
>>
>>
>> --
>>
>> Gareth Bult
>> "The odds of hitting your target go up dramatically when you aim at it."
>> See the status of my current project at http://vdc-store.com
>>
>>
>> - Original Message -
>> From: "Stefan Kooman" 
>> To: "Javier Fontan" 
>> Cc: "Gareth Bult" , users@lists.opennebula.org
>> Sent: Thursday, 30 January, 2014 2:38:23 PM
>> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>>
>> Quoting Javier Fontan (jfon...@opennebula.org):
>>> I've been digging a bit more in the shareable issue and it seems that
>>> setting "shareable" to a disk just disables cache in qemu/kvm[1]:
>>>
>>> --8<--
>>> 4158 virBufferAsprintf(&opt, ",cache=%s", mode);
>>> 4159 } else if (disk->shared && !disk->readonly) {
>>> 4160 virBufferAddLit(&opt, ",cache=off");
>>> 4161 }
>>> -->8--
>>>
>>> Isn't it enough to disable cache for disks that can be migrated?
>> Yes it is, at least in our environment (Ubuntu 13.10). Cache has to be
>> off (to avoid data loss while migrating vm).
>>
>> /etc/one/vmm_exec/vmm_exec_kvm.conf:
>>
>> DISK = [ driver = "raw" , cache = "none", io = "native"]
>>
>> Gr. Stefan
>>
>>
>> --
>> | BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
>> | GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
>
>
>
> --
> Javier Fontán Muiños
> Developer
> OpenNebula - The Open Source Toolkit for Data Center Virtualization
> www.OpenNebula.org | @OpenNebula | github.com/jfontan



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Gareth Bult]

Mmm, I can't disagree if that's what the code says .. however turning
off the cache completely in the GUI does "seem" to make quite a *dramatic*
difference to the VM performance, whereas setting the sharable flag seems
to have no performance impact at all .. (when using writethrough ...)  ???

-- 

Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 
See the status of my current project at http://vdc-store.com 


- Original Message -
From: "Javier Fontan" 
To: "Gareth Bult" 
Cc: "Stefan Kooman" , users@lists.opennebula.org
Sent: Thursday, 30 January, 2014 2:48:15 PM
Subject: Re: [one-users] Setting the "SHAREABLE" attribute

I'll investigate a bit more on this issue but from the qemu driver
code it seems that cache is disabled with shareable flag.

On Thu, Jan 30, 2014 at 4:07 PM, Gareth Bult  wrote:
> Mmm, there has been some discussion around this ..
>
> Technically using cache=writeback "should" be safe for migration.
> (which is the method I use)
>
> So .. I need "cache=writeback" and "sharable" for migration to happen.
> (the performance hit with cache=off is an unacceptable performance hit, at 
> least for me ..)
>
> From the QEMU lists, 2012;
> "In short, if you're using a recent kernel with ext3 or ext4, cache=writeback 
> is absolutely safe. If you're using an older version of ext3, cache=writeback 
> is still safe but ext3 itself isn't. cache=writeback can make the situation 
> worse."
>
> ...
>
>
> --
>
> Gareth Bult
> "The odds of hitting your target go up dramatically when you aim at it."
> See the status of my current project at http://vdc-store.com
>
>
> - Original Message -
> From: "Stefan Kooman" 
> To: "Javier Fontan" 
> Cc: "Gareth Bult" , users@lists.opennebula.org
> Sent: Thursday, 30 January, 2014 2:38:23 PM
> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>
> Quoting Javier Fontan (jfon...@opennebula.org):
>> I've been digging a bit more in the shareable issue and it seems that
>> setting "shareable" to a disk just disables cache in qemu/kvm[1]:
>>
>> --8<--
>> 4158 virBufferAsprintf(&opt, ",cache=%s", mode);
>> 4159 } else if (disk->shared && !disk->readonly) {
>> 4160 virBufferAddLit(&opt, ",cache=off");
>> 4161 }
>> -->8--
>>
>> Isn't it enough to disable cache for disks that can be migrated?
> Yes it is, at least in our environment (Ubuntu 13.10). Cache has to be
> off (to avoid data loss while migrating vm).
>
> /etc/one/vmm_exec/vmm_exec_kvm.conf:
>
> DISK = [ driver = "raw" , cache = "none", io = "native"]
>
> Gr. Stefan
>
>
> --
> | BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
> | GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Javier Fontan]

I'll investigate a bit more on this issue but from the qemu driver
code it seems that cache is disabled with shareable flag.

On Thu, Jan 30, 2014 at 4:07 PM, Gareth Bult  wrote:
> Mmm, there has been some discussion around this ..
>
> Technically using cache=writeback "should" be safe for migration.
> (which is the method I use)
>
> So .. I need "cache=writeback" and "sharable" for migration to happen.
> (the performance hit with cache=off is an unacceptable performance hit, at 
> least for me ..)
>
> From the QEMU lists, 2012;
> "In short, if you're using a recent kernel with ext3 or ext4, cache=writeback 
> is absolutely safe. If you're using an older version of ext3, cache=writeback 
> is still safe but ext3 itself isn't. cache=writeback can make the situation 
> worse."
>
> ...
>
>
> --
>
> Gareth Bult
> "The odds of hitting your target go up dramatically when you aim at it."
> See the status of my current project at http://vdc-store.com
>
>
> - Original Message -
> From: "Stefan Kooman" 
> To: "Javier Fontan" 
> Cc: "Gareth Bult" , users@lists.opennebula.org
> Sent: Thursday, 30 January, 2014 2:38:23 PM
> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>
> Quoting Javier Fontan (jfon...@opennebula.org):
>> I've been digging a bit more in the shareable issue and it seems that
>> setting "shareable" to a disk just disables cache in qemu/kvm[1]:
>>
>> --8<--
>> 4158 virBufferAsprintf(&opt, ",cache=%s", mode);
>> 4159 } else if (disk->shared && !disk->readonly) {
>> 4160 virBufferAddLit(&opt, ",cache=off");
>> 4161 }
>> -->8--
>>
>> Isn't it enough to disable cache for disks that can be migrated?
> Yes it is, at least in our environment (Ubuntu 13.10). Cache has to be
> off (to avoid data loss while migrating vm).
>
> /etc/one/vmm_exec/vmm_exec_kvm.conf:
>
> DISK = [ driver = "raw" , cache = "none", io = "native"]
>
> Gr. Stefan
>
>
> --
> | BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
> | GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Gareth Bult]

Mmm, there has been some discussion around this ..

Technically using cache=writeback "should" be safe for migration.
(which is the method I use)

So .. I need "cache=writeback" and "sharable" for migration to happen.
(the performance hit with cache=off is an unacceptable performance hit, at 
least for me ..)

From the QEMU lists, 2012;
"In short, if you're using a recent kernel with ext3 or ext4, cache=writeback 
is absolutely safe. If you're using an older version of ext3, cache=writeback 
is still safe but ext3 itself isn't. cache=writeback can make the situation 
worse."

...


-- 

Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 
See the status of my current project at http://vdc-store.com 


- Original Message -
From: "Stefan Kooman" 
To: "Javier Fontan" 
Cc: "Gareth Bult" , users@lists.opennebula.org
Sent: Thursday, 30 January, 2014 2:38:23 PM
Subject: Re: [one-users] Setting the "SHAREABLE" attribute

Quoting Javier Fontan (jfon...@opennebula.org):
> I've been digging a bit more in the shareable issue and it seems that
> setting "shareable" to a disk just disables cache in qemu/kvm[1]:
> 
> --8<--
> 4158 virBufferAsprintf(&opt, ",cache=%s", mode);
> 4159 } else if (disk->shared && !disk->readonly) {
> 4160 virBufferAddLit(&opt, ",cache=off");
> 4161 }
> -->8--
> 
> Isn't it enough to disable cache for disks that can be migrated?
Yes it is, at least in our environment (Ubuntu 13.10). Cache has to be
off (to avoid data loss while migrating vm).

/etc/one/vmm_exec/vmm_exec_kvm.conf:

DISK = [ driver = "raw" , cache = "none", io = "native"]

Gr. Stefan
 

-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Stefan Kooman]

Quoting Javier Fontan (jfon...@opennebula.org):
> I've been digging a bit more in the shareable issue and it seems that
> setting "shareable" to a disk just disables cache in qemu/kvm[1]:
> 
> --8<--
> 4158 virBufferAsprintf(&opt, ",cache=%s", mode);
> 4159 } else if (disk->shared && !disk->readonly) {
> 4160 virBufferAddLit(&opt, ",cache=off");
> 4161 }
> -->8--
> 
> Isn't it enough to disable cache for disks that can be migrated?
Yes it is, at least in our environment (Ubuntu 13.10). Cache has to be
off (to avoid data loss while migrating vm).

/etc/one/vmm_exec/vmm_exec_kvm.conf:

DISK = [ driver = "raw" , cache = "none", io = "native"]

Gr. Stefan
 

-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Sunstone Cloud View -- Networks?

[Campbell, Bill]

One step ahead of you, request already submitted.  ;-)

http://dev.opennebula.org/issues/2687

Thanks

-Original Message-
From: Tino Vazquez [mailto:cvazq...@c12g.com]
Sent: Thursday, January 30, 2014 6:03 AM
To: Campbell, Bill
Cc: users
Subject: Re: [one-users] Sunstone Cloud View -- Networks?

Hi Bill,

That is an interesting feature, could you please open a ticket request in
dev.opennebula.org? This way we will be able to schedule it within our
roadmap.

Ideally, this would be configurable, there are scenarios where admins
would want users *not* to change the network of the VM template.

Regards,

-Tino

--
OpenNebula - Flexible Enterprise Cloud Made Simple

--
Constantino Vázquez Blanco, PhD, MSc
Senior Infrastructure Architect at C12G Labs www.c12g.com | @C12G |
es.linkedin.com/in/tinova

--
Confidentiality Warning: The information contained in this e-mail and any
accompanying documents, unless otherwise expressly indicated, is
confidential and privileged, and is intended solely for the person and/or
entity to whom it is addressed (i.e. those identified in the "To" and "cc"
box). They are the property of C12G Labs S.L..
Unauthorized distribution, review, use, disclosure, or copying of this
communication, or any part thereof, is strictly prohibited and may be
unlawful. If you have received this e-mail in error, please notify us
immediately by e-mail at ab...@c12g.com and delete the e-mail and
attachments and any copy from your system. C12G thanks you for your
cooperation.


On 23 January 2014 16:36, Campbell, Bill 
wrote:
> Hello,
>
> I really like the new Cloud sunstone view, but was wondering if it was
> possible to add in the ability to select the network the instance will
> connect to, similar to how the image is selectable?
>
>
>
> 
>
> Bill Campbell
> Infrastructure Architect
>
>
> Axcess Financial Services, Inc.
> 7755 Montgomery Rd., Suite 400
> Cincinnati, OH  45236
>
>
> NOTICE: Protect the information in this message in accordance with the
> company's security policies. If you received this message in error,
> immediately notify the sender and destroy all copies.
>
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
NOTICE: Protect the information in this message in accordance with the 
company's security policies. If you received this message in error, immediately 
notify the sender and destroy all copies.
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Gareth Bult]

Mmm, I'm not entirely sure "which" cache is turned off here, but the version of 
libvirt I seem to be running refuses to migrate an attached block device unless 
it's
marked as sharable. It's not that it fails physically, libvirt just refuses to 
do it ...

(I struggled for quite some time with being unable to migrate until I came 
across this ...)

[Ubuntu server 13.10]

-- 

Gareth Bult 
“The odds of hitting your target go up dramatically when you aim at it.” 
See the status of my current project at http://vdc-store.com 


- Original Message -
From: "Javier Fontan" 
To: "Gareth Bult" 
Cc: users@lists.opennebula.org
Sent: Thursday, 30 January, 2014 12:12:03 PM
Subject: Re: [one-users] Setting the "SHAREABLE" attribute

I've been digging a bit more in the shareable issue and it seems that
setting "shareable" to a disk just disables cache in qemu/kvm[1]:

--8<--
4158 virBufferAsprintf(&opt, ",cache=%s", mode);
4159 } else if (disk->shared && !disk->readonly) {
4160 virBufferAddLit(&opt, ",cache=off");
4161 }
-->8--

Isn't it enough to disable cache for disks that can be migrated?

[1] 
http://libvirt.org/git/?p=libvirt.git;a=blob;f=src/qemu/qemu_command.c;h=e1ff2872209de43a5d9591d7096ca887ed429df3;hb=HEAD#l4159

On Tue, Nov 19, 2013 at 1:16 PM, Gareth Bult  wrote:
> Hi,
>
> I've bypassed this for now by adding a line to "deploy" that says;
> /var/lib/one/remotes/hooks/vdc/deploy.py $domain
> #(just below cat > $domain)
>
> And adding the python script;
>
> #!/usr/bin/python
>
> import xml.etree.ElementTree as ET
> from sys import argv
> original = ET.parse(argv[1])
> params = original.findall(".//disk")
> for p in params: if p.get('device') == 'disk':
> p.insert(0,ET.Element("shareable",{}))
> original.write(argv[1])
>
> This makes all devices of type "disk" shareable.
>
> I *think* anyone using current versions of KVM/libvirt who want to do live
> migration will need this ... unless
> they set "cache=none", which doesn't always do great things for your
> performance ...
>
>
> --
> Gareth Bult
> "The odds of hitting your target go up dramatically when you aim at it."
>
>
>
> 
> From: "Javier Fontan" 
> To: "Gareth Bult" 
> Cc: users@lists.opennebula.org
> Sent: Tuesday, 19 November, 2013 11:54:07 AM
> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>
>
> Right now there is no way to set this value as the deployment file (libvirt
> xml file) is generated in the core and lacks this functionality.
>
> We are evaluating the best way to fix. Adding a "raw" parameter for DISKS
> will be enough or would it be better to have also a default RAW section for
> all the disks?
>
> There are more generic and maybe better ways to add this feature but we are
> really near to the 4.4 release and we are not sure if they can make the
> feature cut.
>
> Cheers
>
>
> On Mon, Nov 18, 2013 at 11:57 AM, Gareth Bult  wrote:
>>
>> Hi,
>>
>> For the most recent versions of libvirt, in order to do a live migration
>> with writethru cache turned
>> on, the "" attribute needs to be set in the DISK section of
>> the configuration. "virt-manager"
>> has a tick box for this in it's disk setup section .. is there an easy way
>> to implement this in OpenNebula?
>>
>> tia
>>
>> --
>> Gareth Bult
>> "The odds of hitting your target go up dramatically when you aim at it."
>>
>>
>>
>>
>> ___
>> Users mailing list
>> Users@lists.opennebula.org
>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>
>
>
>
> --
> Javier Fontán Muiños
> Developer
> OpenNebula - The Open Source Toolkit for Data Center Virtualization
> www.OpenNebula.org | @OpenNebula | github.com/jfontan
>



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Setting the "SHAREABLE" attribute

[Javier Fontan]

I've been digging a bit more in the shareable issue and it seems that
setting "shareable" to a disk just disables cache in qemu/kvm[1]:

--8<--
4158 virBufferAsprintf(&opt, ",cache=%s", mode);
4159 } else if (disk->shared && !disk->readonly) {
4160 virBufferAddLit(&opt, ",cache=off");
4161 }
-->8--

Isn't it enough to disable cache for disks that can be migrated?

[1] 
http://libvirt.org/git/?p=libvirt.git;a=blob;f=src/qemu/qemu_command.c;h=e1ff2872209de43a5d9591d7096ca887ed429df3;hb=HEAD#l4159

On Tue, Nov 19, 2013 at 1:16 PM, Gareth Bult  wrote:
> Hi,
>
> I've bypassed this for now by adding a line to "deploy" that says;
> /var/lib/one/remotes/hooks/vdc/deploy.py $domain
> #(just below cat > $domain)
>
> And adding the python script;
>
> #!/usr/bin/python
>
> import xml.etree.ElementTree as ET
> from sys import argv
> original = ET.parse(argv[1])
> params = original.findall(".//disk")
> for p in params: if p.get('device') == 'disk':
> p.insert(0,ET.Element("shareable",{}))
> original.write(argv[1])
>
> This makes all devices of type "disk" shareable.
>
> I *think* anyone using current versions of KVM/libvirt who want to do live
> migration will need this ... unless
> they set "cache=none", which doesn't always do great things for your
> performance ...
>
>
> --
> Gareth Bult
> "The odds of hitting your target go up dramatically when you aim at it."
>
>
>
> 
> From: "Javier Fontan" 
> To: "Gareth Bult" 
> Cc: users@lists.opennebula.org
> Sent: Tuesday, 19 November, 2013 11:54:07 AM
> Subject: Re: [one-users] Setting the "SHAREABLE" attribute
>
>
> Right now there is no way to set this value as the deployment file (libvirt
> xml file) is generated in the core and lacks this functionality.
>
> We are evaluating the best way to fix. Adding a "raw" parameter for DISKS
> will be enough or would it be better to have also a default RAW section for
> all the disks?
>
> There are more generic and maybe better ways to add this feature but we are
> really near to the 4.4 release and we are not sure if they can make the
> feature cut.
>
> Cheers
>
>
> On Mon, Nov 18, 2013 at 11:57 AM, Gareth Bult  wrote:
>>
>> Hi,
>>
>> For the most recent versions of libvirt, in order to do a live migration
>> with writethru cache turned
>> on, the "" attribute needs to be set in the DISK section of
>> the configuration. "virt-manager"
>> has a tick box for this in it's disk setup section .. is there an easy way
>> to implement this in OpenNebula?
>>
>> tia
>>
>> --
>> Gareth Bult
>> "The odds of hitting your target go up dramatically when you aim at it."
>>
>>
>>
>>
>> ___
>> Users mailing list
>> Users@lists.opennebula.org
>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>
>
>
>
> --
> Javier Fontán Muiños
> Developer
> OpenNebula - The Open Source Toolkit for Data Center Virtualization
> www.OpenNebula.org | @OpenNebula | github.com/jfontan
>



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Sunstone Cloud View -- Networks?

[Tino Vazquez]

Hi Bill,

That is an interesting feature, could you please open a ticket request
in dev.opennebula.org? This way we will be able to schedule it within
our roadmap.

Ideally, this would be configurable, there are scenarios where admins
would want users *not* to change the network of the VM template.

Regards,

-Tino

--
OpenNebula - Flexible Enterprise Cloud Made Simple

--
Constantino Vázquez Blanco, PhD, MSc
Senior Infrastructure Architect at C12G Labs
www.c12g.com | @C12G | es.linkedin.com/in/tinova

--
Confidentiality Warning: The information contained in this e-mail and
any accompanying documents, unless otherwise expressly indicated, is
confidential and privileged, and is intended solely for the person
and/or entity to whom it is addressed (i.e. those identified in the
"To" and "cc" box). They are the property of C12G Labs S.L..
Unauthorized distribution, review, use, disclosure, or copying of this
communication, or any part thereof, is strictly prohibited and may be
unlawful. If you have received this e-mail in error, please notify us
immediately by e-mail at ab...@c12g.com and delete the e-mail and
attachments and any copy from your system. C12G thanks you for your
cooperation.


On 23 January 2014 16:36, Campbell, Bill  wrote:
> Hello,
>
> I really like the new Cloud sunstone view, but was wondering if it was
> possible to add in the ability to select the network the instance will
> connect to, similar to how the image is selectable?
>
>
>
> 
>
> Bill Campbell
> Infrastructure Architect
>
>
> Axcess Financial Services, Inc.
> 7755 Montgomery Rd., Suite 400
> Cincinnati, OH  45236
>
>
> NOTICE: Protect the information in this message in accordance with the
> company's security policies. If you received this message in error,
> immediately notify the sender and destroy all copies.
>
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] External snapshots

[Carlos Martín Sánchez]

Hi,

OpenNebula does not manage kvm external snapshots. You could modify the
snapshot scripts [1] and then schedule onevm snapshot-create actions. But
this may have side effects, since OpenNebula won't be aware of any other
new VM files.

If you modify the drivers in the front-end's /var/lib/one/remotes dir,
you'll need to execute onehost sync in order to propagate the changes to
the hosts.

Regards

[1]
https://github.com/OpenNebula/one/blob/master/src/vmm_mad/remotes/kvm/snapshot_create

--
Carlos Martín, MSc
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | cmar...@opennebula.org |
@OpenNebula


On Mon, Jan 27, 2014 at 5:13 PM, Giancarlo De Filippis
wrote:

>  Hi all,
>
> Is possible to configure scheduled external snapshots from sunstone or i
> must create script for use virsh directly (stop VM and create snapshots)?
>
> Thank all for answer...
>
> Giancarlo
>
>
>
> PS I've glusterfs datastore...
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] HA hook and "-p " flag

[Tino Vazquez]

Hi Igor,

Just to let you know, this has been fixed in the main repo, and will
be available for upcoming releases (4.4.1 and 4.6).

Regards,

-Tino

--
OpenNebula - Flexible Enterprise Cloud Made Simple

--
Constantino Vázquez Blanco, PhD, MSc
Senior Infrastructure Architect at C12G Labs
www.c12g.com | @C12G | es.linkedin.com/in/tinova

--
Confidentiality Warning: The information contained in this e-mail and
any accompanying documents, unless otherwise expressly indicated, is
confidential and privileged, and is intended solely for the person
and/or entity to whom it is addressed (i.e. those identified in the
"To" and "cc" box). They are the property of C12G Labs S.L..
Unauthorized distribution, review, use, disclosure, or copying of this
communication, or any part thereof, is strictly prohibited and may be
unlawful. If you have received this e-mail in error, please notify us
immediately by e-mail at ab...@c12g.com and delete the e-mail and
attachments and any copy from your system. C12G thanks you for your
cooperation.


On 29 January 2014 16:42, Tino Vazquez  wrote:
> Hi Igor,
>
> Awesome, thanks a lot for letting us know.
>
> Best,
>
> -Tino
>
> --
> OpenNebula - Flexible Enterprise Cloud Made Simple
>
> --
> Constantino Vázquez Blanco, PhD, MSc
> Senior Infrastructure Architect at C12G Labs
> www.c12g.com | @C12G | es.linkedin.com/in/tinova
>
> --
> Confidentiality Warning: The information contained in this e-mail and
> any accompanying documents, unless otherwise expressly indicated, is
> confidential and privileged, and is intended solely for the person
> and/or entity to whom it is addressed (i.e. those identified in the
> "To" and "cc" box). They are the property of C12G Labs S.L..
> Unauthorized distribution, review, use, disclosure, or copying of this
> communication, or any part thereof, is strictly prohibited and may be
> unlawful. If you have received this e-mail in error, please notify us
> immediately by e-mail at ab...@c12g.com and delete the e-mail and
> attachments and any copy from your system. C12G thanks you for your
> cooperation.
>
>
> On 29 January 2014 16:41, Igor Laskovy  wrote:
>> Hi Tino,
>>
>> Thanks, yes and sorry for delay.
>> I would like confirm for now that with this change it's works.
>>
>>
>> On Wed, Jan 29, 2014 at 7:38 PM, Tino Vazquez  wrote:
>>>
>>> Hi Igor,
>>>
>>> I've opened a ticket [1] to keep track of the solution for next releases.
>>>
>>> Thanks for the feedback,
>>>
>>> -Tino
>>>
>>> [1] http://dev.opennebula.org/issues/2689
>>> --
>>> OpenNebula - Flexible Enterprise Cloud Made Simple
>>>
>>> --
>>> Constantino Vázquez Blanco, PhD, MSc
>>> Senior Infrastructure Architect at C12G Labs
>>> www.c12g.com | @C12G | es.linkedin.com/in/tinova
>>>
>>> --
>>> Confidentiality Warning: The information contained in this e-mail and
>>> any accompanying documents, unless otherwise expressly indicated, is
>>> confidential and privileged, and is intended solely for the person
>>> and/or entity to whom it is addressed (i.e. those identified in the
>>> "To" and "cc" box). They are the property of C12G Labs S.L..
>>> Unauthorized distribution, review, use, disclosure, or copying of this
>>> communication, or any part thereof, is strictly prohibited and may be
>>> unlawful. If you have received this e-mail in error, please notify us
>>> immediately by e-mail at ab...@c12g.com and delete the e-mail and
>>> attachments and any copy from your system. C12G thanks you for your
>>> cooperation.
>>>
>>>
>>> On 24 January 2014 12:00, Tino Vazquez  wrote:
>>> > Dear Igor,
>>> >
>>> > This looks like a bug indeed. Could you try changing
>>> > /var/lib/one/remotes//hooks/ft/host_error.rb, line 98, from:
>>> >
>>> >  monitor_interval = line.split("=").last.to_i if
>>> > /HOST_MONITORING_INTERVAL/=~line
>>> >
>>> > to
>>> >
>>> > monitor_interval = line.split("=").last.to_i if
>>> > /MONITORING_INTERVAL/=~line
>>> >
>>> > and try again. If this works, please let us know so we can include a fix
>>> > for 4.6
>>> >
>>> > Regards,
>>> >
>>> > -Tino
>>> >
>>> > --
>>> > OpenNebula - Flexible Enterprise Cloud Made Simple
>>> >
>>> > --
>>> > Constantino Vázquez Blanco, PhD, MSc
>>> > Senior Infrastructure Architect at C12G Labs
>>> > www.c12g.com | @C12G | es.linkedin.com/in/tinova
>>> >
>>> > --
>>> > Confidentiality Warning: The information contained in this e-mail and
>>> > any accompanying documents, unless otherwise expressly indicated, is
>>> > confidential and privileged, and is intended solely for the person
>>> > and/or entity to whom it is addressed (i.e. those identified in the
>>> > "To" and "cc" box). They are the property of C12G Labs S.L..
>>> > Unauthorized distribution, review, use, disclosure, or copying of this
>>> > communication, or any part thereof, is strictly prohibited and may be
>>> > unlawful. If you have received this e-mail in error, please notify us
>>> > immediately by e-mail at ab...@c12g.com and delete the e-mail and
>>> > attachments and any copy

Re: [one-users] Vlan template

[Jaime Melis]

Hi all,

Stefan Kooman recently requested this functionality and it's included in
the code. Apologies for not understanding the issue better before:

http://dev.opennebula.org/issues/2345

cheers,
Jaime


On Tue, Jan 28, 2014 at 9:42 PM, Dmitri Chebotarov  wrote:

> It seems like the command to add trunk is almost identical to adding vlan:
>
> - vlan (ie. Access-port) : ovs-vsctl set port vnet0 tag=Vlan1
> - trunk : ovs-vsctl set port vnet0 trunks=VlanID1,VlanID2
>
> Would it be possible to adjust command to add trunk port with list of
> VLANs if multiple VLAN Ids listed in the Virtual Network’s VLAN field?
>
> --
> Thank you,
>
> Dmitri Chebotarov
> VCL Sys Eng, Engineering & Architectural Support, TSD - Ent Servers &
> Messaging
> 223 Aquia Building, Ffx, MSN: 1B5
> Phone: (703) 993-6175 | Fax: (703) 993-3404
>
>
>
>
>
>
>
> On 1/28/14, 12:23 , "Dmitri Chebotarov"  wrote:
>
> >Hi guys
> >
> >Have you found solution for how to add trunk interface?
> >
> >I¹m facing the same task and looking for a possible solution. I found this
> >ports
> >
> http://blog.scottlowe.org/2013/05/28/vlan-trunking-to-guest-domains-with-o
> >p
> >en-vswitch/ and trying to implement it via RAW entry in VM template. No
> >results yetŠ
> >
> >--
> >Thank you,
> >
> >Dmitri Chebotarov
> >VCL Sys Eng, Engineering & Architectural Support, TSD - Ent Servers &
> >Messaging
> >223 Aquia Building, Ffx, MSN: 1B5
> >Phone: (703) 993-6175 | Fax: (703) 993-3404
> >
> >
> >
> >
> >
> >
> >
> >On 5/21/13, 3:21 , "Gandalf Corvotempesta"
> > wrote:
> >
> >>2013/5/20 Jaime Melis :
> >>> However, you can only tag one VLAN per interfaces, which as far as I
> >>>know,
> >>> is the only thing that makes sense?
> >>
> >>Why? Trunk interfaces carries multiple vlan, is absolutely standard.
> >>I have not less than 58 VLANs on a single interface on my XenServer host.
> >>___
> >>Users mailing list
> >>Users@lists.opennebula.org
> >>http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
> >>
> >
> >___
> >Users mailing list
> >Users@lists.opennebula.org
> >http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>



-- 
Jaime Melis
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | jme...@opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Using CloudInit for contextualization

[Javier Fontan]

Unfortunately the current version of cloud-init does not load new
network parameters after they are configured in some distributions.
There is a ticket to track that problem [1]

The documentation gives some ideas on how to overcome this [2]:

--8<--
The current version of cloud-init configures the network before
running cloud-init configuration. This makes the network configuration
not reliable. Until a new version that fixes this is released you can
add OpenNebula context packages or this user data to reboot the
machine so the network is properly configured.
-->8--

[1] https://bugs.launchpad.net/cloud-init/+bug/1225922
[2] http://docs.opennebula.org/stable/user/virtual_machine_setup/cloud-init.html

On Tue, Jan 28, 2014 at 1:56 PM, Wilma Hermann  wrote:
> Hi,
>
> I'm trying to get a Ubuntu 13.10 guest to work, which uses a official
> Cloud-image (http://cloud-images.ubuntu.com/saucy/) as disk. However, the VM
> is not integrating into the virtual network. I've double checked the virtual
> network settings with CentOS-VMs, the network works correctly.
>
> I've found out that CloudInit writes a correct /etc/network/interfaces file
> with a static address, however, the machine somehow ignores this file and
> keeps sending DHCP requests without getting an answer (which is intended
> since we don't use DHCP in the VM's network).
>
> # virt-cat one-42 /etc/network/interfaces
> auto lo
> iface lo inet loopback
>
> auto eth0
> iface eth0 inet static
>   address 192.168.129.4
>   network 192.168.128.0
>   netmask 255.255.254.0
>   gateway 192.168.128.1
>
> # virt-cat one-42 /var/log/syslog | tail -6
> Jan 28 12:42:44 ubuntu dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
> port 67 interval 21 (xid=0x1080a577)
> Jan 28 12:43:05 ubuntu dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
> port 67 interval 13 (xid=0x1080a577)
> Jan 28 12:43:18 ubuntu dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
> port 67 interval 20 (xid=0x1080a577)
> Jan 28 12:43:38 ubuntu dhclient: DHCPDISCOVER on eth0 to 255.255.255.255
> port 67 interval 1 (xid=0x1080a577)
> Jan 28 12:43:39 ubuntu dhclient: No DHCPOFFERS received.
> Jan 28 12:43:39 ubuntu dhclient: No working leases in persistent database -
> sleeping.
>
> Does anybody know how the VM template needs to be configured in order to get
> it working?
>
> Thanks in advance
> Wilma
>
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>



-- 
Javier Fontán Muiños
Developer
OpenNebula - The Open Source Toolkit for Data Center Virtualization
www.OpenNebula.org | @OpenNebula | github.com/jfontan
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] [SOLVED] libvirt/kvm: Permission denied on PERSISTENT images

[Jaime Melis]

Hi Daniel,

thanks for the update. We've seen the *bump*  to the associated issue, and
we agree :)
http://dev.opennebula.org/issues/2381#change-7555

cheers,
Jaime


On Wed, Jan 29, 2014 at 4:36 PM, Daniel Dehennin <
daniel.dehen...@baby-gnu.org> wrote:

> Daniel Dehennin  writes:
>
> > Hello, again,
> >
> > The problem is libvirt specific:
>
> [...]
>
> > I can do an "md5sum /var/lib/one//datastores/0/1937/disk.0".
>
> As "oneadmin", but libvirtd is running as root and new version make some
> checks on the file, which failed because of a "root_squash"[1]
>
> Changing the exportfs parameter solve my issue, sorry for the noises.
>
> Footnotes:
> [1]  http://dev.opennebula.org/issues/2381
>
> --
> Daniel Dehennin
> Récupérer ma clef GPG:
> gpg --keyserver pgp.mit.edu --recv-keys 0x7A6FE2DF
>
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>


-- 
Jaime Melis
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | jme...@opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] Unable to auto start one and sunstone service in Ubuntu 13.10

[Jaime Melis]

Hi Kiran,

That sounds very interesting, at some point it would be great if you could
elaborate a bit more on exactly how you have done it, but it looks really
nice.

The thing about self contained mode is that it's desgined for development
or for custom solutions, such as yours. It's not intented for packaging or
to provide commercial supoprt.

So, in your particular case, I think you can fix this quite easily.
Continue with the self-contained mode, but instead of copying those scripts
(one and sunstone-server), install these files to /etc/init.d:
https://github.com/OpenNebula/one/blob/master/share/pkgs/Ubuntu/opennebula
https://github.com/OpenNebula/one/blob/master/share/pkgs/Ubuntu/opennebula-sunstone

Then, make some changes to them so they work for system contained mode. It
shouldn't be much change, maybe just sourcing the file were you have the
ONE_LOCATION variable defined before doing one start and one stop.

Let us know how it goes.

Cheers,
Jaime






On Mon, Jan 27, 2014 at 8:06 PM, kiran ranjane wrote:

> Hi Jaime,
>
> May I know which version of Ubuntu is officially supported for self
> contained mode as we will be only using self-contain mode for Opennebula
> because we have setup Automatic failover and failback of cloud management
> without using luci/ricci or pacemaker/corosync.
>
> This setup is simple and very robust using ucarp,rsync and master-master
> mysql replication, easy to setup and maintain, no complex things, no split
> brain and works like a charm.
>
> Thanks and Regards
> Kiran Ranjane
>
>
> On Mon, Jan 27, 2014 at 3:39 PM, Jaime Melis  wrote:
>
>> Hi,
>>
>> 13.10 is not yet supported. Can you instead use 13.04? You would be able
>> to use packages then.
>>
>> If you must use 13.10, you should install from source in *system-wide*
>> mode, not self-contained, and the copy these scripts to /etc/init.d (no the
>> ones in the bin folder):
>> https://github.com/OpenNebula/one/tree/master/share/pkgs/Ubuntu
>>
>> regards,
>> Jaime
>>
>>
>> On Fri, Jan 24, 2014 at 1:25 PM, kiran ranjane 
>> wrote:
>>
>>> Hello,
>>>
>>> I have installed ONE 4.4 from source and it is working fine, However
>>> After reboot of the server one and sunstone service does not start, I have
>>> to start it manually.
>>>
>>> -- I have copied the one and sunstone script from the bin folder to
>>> /etc/init.d/ and also updated it in rc using update-rc.d one defaults and
>>> same for sunstone but it still does not start after a reboot.
>>>
>>> -- I have install 4.4 in self-contained mode and variables are properly
>>> configured as I can get the one command from shell prompt.
>>>
>>> Kindly let me know what else I can do to start the service after a reboot
>>>
>>> Thanks and Regards
>>> Kiran Ranjane
>>>
>>> ___
>>> Users mailing list
>>> Users@lists.opennebula.org
>>> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>>>
>>>
>>
>


-- 
Jaime Melis
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | jme...@opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] CentOS Dojo Hackathon

[Jaime Melis]

Hi Steven,

I don't think there are any news with regard to that as of yet, but as soon
as there is, we'll definitely share it with the list.

Regards,
Jaime


On Wed, Jan 29, 2014 at 3:24 PM, Steven Timm  wrote:

>
> For those of us who can't make it to Brussels, it would be interesting
> to hear if there have been any changes in OpenNebula on CentOS that
> have come about because of the new affiliation of Centos and Red Hat.
> In particular, might there now be a path to get OpenNebula and some
> of the other packages that have been CentOS-only in to Red Hat Enterprise
> Linux?
>
> Steve Timm
>
>
> On Wed, 29 Jan 2014, Jaime Melis wrote:
>
>  Dear all,
>> as I mentioned some days ago in the list, OpenNebula will participate in
>> the
>> upcoming CentOS Dojo:
>> http://wiki.centos.org/Events/Dojo/Brussels2014
>>
>> We have a bit more information with regard to this. The hackathon will
>> start
>> shortly after lunch, at 14:15, and we will take upon us to improve the
>> following points in OpenNebula / CentOS:
>>
>> - CloudInit 0.7.5 support
>> - Systemd scripts for CentOS 7
>> - opennebula-node-xen package
>>
>> If you are interested please join us (even remotely!), any help will be
>> highly appreciated!
>>
>> cheers,
>> Jaime
>>
>> --
>> Jaime Melis
>> Project Engineer
>> OpenNebula - Flexible Enterprise Cloud Made Simple
>> www.OpenNebula.org | jme...@opennebula.org
>>
>>
>>
> --
> Steven C. Timm, Ph.D  (630) 840-8525
> t...@fnal.gov  http://home.fnal.gov/~timm/
> Fermilab Scientific Computing Division, Scientific Computing Services Quad.
> Grid and Cloud Services Dept., Associate Dept. Head for Cloud Computing
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>
>


-- 
Jaime Melis
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | jme...@opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] CentOS Dojo Hackathon

[Jaime Melis]

Hi Stefan,

I agree - I'm actually writing this email from Arch Linux, actually :-).

I don't have much systemd expertise, but from looking at some examples it
looks something that we could easily do. Let's see how systemd's
documentation is, as long as it's clear enough...

Robert Schweikert made some custom systemd scripts for openSUSE, we can
base our work off of that. The good thing about OpenNebula is that it has
very few distro-specific patches or tweaks, so I think we can make the
systemd support distro independent. I'm also looking at reusing those
scripts for Fedora.

cheers,
Jaime


On Thu, Jan 30, 2014 at 10:09 AM, Stefan Kooman  wrote:

> Quoting Jaime Melis (jme...@opennebula.org):
> >
> > We have a bit more information with regard to this. The hackathon will
> > start shortly after lunch, at 14:15, and we will take upon us to improve
> > the following points in OpenNebula / CentOS:
> >
> > - CloudInit 0.7.5 support
> > - Systemd scripts for CentOS 7
> It would be cool though, if easily possible, to make systemd support distro
> independent (I'm thinking about Arch Linux).
>
> Gr. Stefan
>
> --
> | BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
> | GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
> ___
> Users mailing list
> Users@lists.opennebula.org
> http://lists.opennebula.org/listinfo.cgi/users-opennebula.org
>



-- 
Jaime Melis
Project Engineer
OpenNebula - Flexible Enterprise Cloud Made Simple
www.OpenNebula.org | jme...@opennebula.org
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

Re: [one-users] CentOS Dojo Hackathon

[Stefan Kooman]

Quoting Jaime Melis (jme...@opennebula.org):
> 
> We have a bit more information with regard to this. The hackathon will
> start shortly after lunch, at 14:15, and we will take upon us to improve
> the following points in OpenNebula / CentOS:
> 
> - CloudInit 0.7.5 support
> - Systemd scripts for CentOS 7
It would be cool though, if easily possible, to make systemd support distro
independent (I'm thinking about Arch Linux).

Gr. Stefan

-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org

[one-users] onevnet policy violation?

[Stefan Kooman]

Hi,

I created a new vnet based on the following template:

--- template start ---
NAME= "v-test01"
TYPE= "ranged"
BRIDGE  = "uplink"
CLUSTER = "kvm_cluster"

NETWORK_ADDRESS = 172.17.226.0
NETWORK_MASK= 255.255.255.224
IP_START= 172.17.226.4
IP_END  = 172.17.226.30

GATEWAY = 172.17.226.1
 
VLAN= "YES"
VLAN_ID = 226
--- template end ---

After creation the vnet is not a member of the kvm_cluster but instead
of the default cluster "-". Is this a bug or is it me doing it wrong?
After reading [1] again it explicitly states a "vnet or datastore"
should be added with the onecluster add command, so I was doing it
wrong. It still would be nice to have it automatically added to a
cluster if "CLUSTER" attribute is there.

After that  I instantiated a vm based on a template with a mix of
resources from "kvm_cluster" and "-" (v-test01). This vm succesfully
managed to get into running state. According to [1] section "The Default
Cluster 'None'" hosts in the Cluster 'none' will only run VMs using
resources without a cluster. I assume this is equally true for hosts
that are a member of a non-default cluster (kvm_cluster in our case) but
instead it did run the vm. Indeed, according to "The "Scheduling and
Cluster" section of [1] the deployment should have failed. What info do
we need to make sure it's a bug and not PEBKAC?

Gr. Stefan

[1]:
http://docs.opennebula.org/4.4/administration/hosts_and_clusters/cluster_guide.html


-- 
| BIT BV  http://www.bit.nl/Kamer van Koophandel 09090351
| GPG: 0xD14839C6   +31 318 648 688 / i...@bit.nl
___
Users mailing list
Users@lists.opennebula.org
http://lists.opennebula.org/listinfo.cgi/users-opennebula.org