Re: [ovirt-users] How to execute Virsh command after ovirt installation ?

[jvandewege]

On 3-1-2017 8:32, TranceWorldLogic . wrote:
> Hi,
>
> I tried execution below virsh command after ovirt installation.
> but it requesting username and password.
>
> ~# virsh list --all
> Please enter your authentication name:
>
> Please let me know which username and password I have to provide ?
If you only need read access then you need to supply -r as an additional
argument.

Why do you need virsh because that will only give you the info on that
host, using ovirt-shell gives you the info for all hosts.

Joop

___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users

[ovirt-users] How to execute Virsh command after ovirt installation ?

[TranceWorldLogic .]

Hi,

I tried execution below virsh command after ovirt installation.
but it requesting username and password.

~# virsh list --all
Please enter your authentication name:

Please let me know which username and password I have to provide ?

Thanks,
~Rohit
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users

Re: [ovirt-users] How to invoke ovirt-guest-agent hook from ovirt engine SDK ?

[TranceWorldLogic .]

Hi,

It look interesting.
Would you please point me some document that cover in detail ?

I was looking for mechanism where I can monitor guest application.
If application get failed or crash then I need to restart or start other VM.
Is it possible in ovirt ?
I though I will use hook but as you explain it clear me that my understand
was wrong.
Thanks for your help.

Thanks,
~Rohit


On Mon, Jan 2, 2017 at 1:04 PM, Vinzenz Feenstra 
wrote:

>
> > On Dec 30, 2016, at 11:03 AM, TranceWorldLogic . <
> tranceworldlo...@gmail.com> wrote:
> >
> > HI,
>
> Hi there,
>
> >
> > I was exploring more about ovirt-gueste-engin.
> > It look to me very easy to configure add add hook as script.
> >
> > But my doubt is, how to invoke those script from ovirt-engine ?
> > Please some one help me to understand this part.
> > I am looking into python SDK code to figure out same but still not got
> luck
>
> Guest agent hooks aren’t triggered through the SDK, hooks are triggered
> when certain events happen on the hypervisor side.
> e.g. A VM gets migrated from HOST A to HOST B or the VM gets suspended. In
> these cases VDSM _can_ send a message to the guest
> agent asking it to process all hooks.
>
> That those hooks are enabled are depending on the migration policy
> configured. Currently all but the ‘Legacy’  migration policies do cause the
> hooks to be executed, given a new enough guest agent, VDSM and cluster
> version.
> HTH
>
>
> >
> > Thanks,
> > ~Rohit
> > ___
> > 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: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network

[Sverker Abrahamsson]

Found an issue with Ovirt - OVN integration.

Engine and OVN central db running on host h2. Created VM to run on host 
h1, which is started. Ovn db state:


[root@h2 env3]# ovn-nbctl show
switch e53554cf-e553-40a1-8d22-9c8d95ec0601 (ovirtbridge)
port 4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873
addresses: ["00:1a:4a:16:01:51"]
port 92f6d3c8-68b3-4986-9c09-60bee04644b5
addresses: ["00:1a:4a:16:01:52"]
port ovirtbridge-port2
addresses: ["unknown"]
port ovirtbridge-port1
addresses: ["unknown"]
[root@h2 env3]# ovn-sbctl show
Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5"
hostname: "h2.limetransit.com"
Encap geneve
ip: "148.251.126.50"
options: {csum="true"}
Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873"
Port_Binding "ovirtbridge-port1"
Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9"
hostname: "h1.limetransit.com"
Encap geneve
ip: "144.76.84.73"
options: {csum="true"}
Port_Binding "ovirtbridge-port2"

Port 92f6d3c8-68b3-4986-9c09-60bee04644b5 is for the new VM which is 
started on h1, but it is not assigned to that chassis. The reason is 
that on h1 the port on br-int is created like this:


ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int 
vnet0 -- set Interface vnet0 
"external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 
"external-ids:iface-id=\"35bcbe31-2c7e-4d97-add9-ce150eeb2f11\"" -- set 
Interface vnet0 
"external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set 
Interface vnet0 external-ids:iface-status=active


I.e. the extrernal id of interface is wrong. When I manually change to 
the right id like this the port works fine:


ovs-vsctl --timeout=5 -- --if-exists del-port vnet0 -- add-port br-int 
vnet0 -- set Interface vnet0 
"external-ids:attached-mac=\"00:1a:4a:16:01:52\"" -- set Interface vnet0 
"external-ids:iface-id=\"92f6d3c8-68b3-4986-9c09-60bee04644b5\"" -- set 
Interface vnet0 
"external-ids:vm-id=\"4d0c134a-11a0-40f4-b2fb-c13c17c7251c\"" -- set 
Interface vnet0 external-ids:iface-status=active


sb db after correcting the port:

Chassis "6e4dd29f-7607-48d7-8e5a-eef4c6aeefb5"
hostname: "h2.limetransit.com"
Encap geneve
ip: "148.251.126.50"
options: {csum="true"}
Port_Binding "4981ee5f-6e15-4bd5-a1cf-7ead9bdd5873"
Port_Binding "ovirtbridge-port1"
Chassis "4f10fb04-8fb2-48d7-8a3f-ea6444c02cf9"
hostname: "h1.limetransit.com"
Encap geneve
ip: "144.76.84.73"
options: {csum="true"}
Port_Binding "ovirtbridge-port2"
Port_Binding "92f6d3c8-68b3-4986-9c09-60bee04644b5"

I don't know from where the faulty id comes from, it's not in any logs. 
In the domain xml as printed in vdsm.log the id is correct:











interfaceid="92f6d3c8-68b3-4986-9c09-60bee04644b5" />




Where is the ovs-vsctl command line built for this call?

/Sverker


Den 2017-01-02 kl. 13:40, skrev Sverker Abrahamsson:
Got it to work now by following the env8 example in OVN tutorial, 
where a port is added with type l2gateway. Not sure how that is 
different from the localnet variant, but didn't suceed in getting that 
one working. Now I'm able to ping and telnet over the tunnel, but not 
ssh even when the port is answering on telnet. Neither does nfs 
traffic work even though mount did. Suspecting MTU issue. I did notice 
that ovn-controller starts too early, before network interfaces are 
established and hence can't reach the db. As these is a purely OVS/OVN 
issue I'll ask about it on their mailing list.


Getting back to the original issue with Ovirt, I've now added the 
second host h1 to ovirt-engine. Had to do the same as with h2 to 
create a dummy ovirtmgmt network but configured access via the public 
IP. My firewall settings was replaced with iptables config and 
vdsm.conf was overwritten when engine was set up, so those had to be 
manually restored. It would be preferable if it would be possible to 
configure ovirt-engine that it does not "own" the host and instead 
comply with the settings it has instead of enforcing it's own view..


Apart from that it seems the second host works, although I need to 
resolve the traffic issue over the OVS tunnel.

/Sverker

Den 2017-01-02 kl. 01:13, skrev Sverker Abrahamsson:
1. That is not possible as ovirt (or vdsm) will rewrite the network 
configuration to a non-working state. That is why I've set that if as 
hidden to vdsm and is why I'm keen on getting OVS/OVN to work


2. I've been reading the doc for OVN and starting to connect the 
dots, which is not trivial as it is complex. Some insights reached:


First step is the OVN database, installed by openvswitch-ovn-central, 
which I currently have running on h2 host. The 'ovn-nbctl' and 
'ovn-sbctl' commands are only possible to execute on a dat

Re: [ovirt-users] Watchdog device

[Gary Pedretty]

I will try it with the vm shutdown.  

thanks

gary



Gary Pedrettyg...@ravnalaska.net
Systems Manager  www.flyravn.com
Ravn Alaska   /\907-450-7251
5245 Airport Industrial Road /  \/\ 907-450-7238 fax
Fairbanks, Alaska  99709/\  /\ \ Second greatest commandment
Serving Alaska's Interior  /  \/  /\  \ \/\   "Love your neighbor as
Having a heatwave, its summer   yourself” Matt 22:39






> On Jan 1, 2017, at 5:18 AM, Doron Fediuck  wrote:
> 
> Hi Gary,
> this is a known issue we're working on. if the VM is down it should
> work as expected (edit the VM and close the dialog).
> Please try and let me know. Another option is to do this until the bug
> is resolved is using the REST API.
> 
> Note that this is currently working on on Linux guests (there are no
> updated Windows drivers). You will need to configure
> the device in the guest as explained by the docs.
> 
> Doron
> 
> 
> On Sun, Jan 1, 2017 at 9:28 AM, Gary Pedretty  wrote:
>> How do you add the supported watchdog device to a VM that has already been
>> created?  I have tried adding it via the High Availability tab in the Edit
>> VM Dialog, but it does not  retain the setting when I close the dialog.
>> 
>> Latest version os Overt running as self-hosted engine
>> 
>> Gary
>> 
>> 
>> 
>> Gary Pedrettyg...@ravnalaska.net
>> Systems Manager  www.flyravn.com
>> Ravn Alaska   /\907-450-7251
>> 5245 Airport Industrial Road /  \/\ 907-450-7238 fax
>> Fairbanks, Alaska  99709/\  /\ \ Second greatest commandment
>> Serving Alaska's Interior  /  \/  /\  \ \/\   "Love your neighbor as
>> Having a heatwave, its summer   yourself” Matt 22:39
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> ___
>> 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

[ovirt-users] import storage domain and import VMs from ovirt 3.6 to 4

[Paul Groeneweg | Pazion]

Hi,

I am finally planning to start our oVirt hosted engine upgrade. I am not
sure how stable my hosted engine storage is ( it shows wrong status in web
GUI ) So I arranged a new shared storage and plan to do a fresh install (
altough I noticed there is also an option to upgrade "hosted-engine
--upgrade-appliance" ).

After I installed the oVirt 4 hosted engine on a host node I want to move
all my running VMs to the new environment. With VMs at first running on the
current data storage domain. My storage domain is FC based.

My idea is to:

1. stop all VMs
2. detach Storage domain on old ovirt 3.6 environment ( and keep data ).
3. import and activate the storage domain on the new ovirt 4 environment.

To prevent any data loss or copy too much data I have the follwoing
questions:

1. Is this storage domain import from 3.6 to 4.0 available?
2. I read I should import all VMs, this means I need to copy all data or is
it just a matter of importing a config, so all VM data /image remains on
the imported storage domain?
3. When for some reason I want to roll back when I have imported the
storage domain without having any VM imported, can I just detach and
re-attach the storage domain to the previous 3.6 environment?

If there might be a better path to get my VMs to a clean oVirt 4.0
environment I am glad to hear.

Thanks in advance!
Kind regards,
Paul
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users

Re: [ovirt-users] New install: can't install engine

[Jim Kusznir]

I did eventually figure out the issue: I mis-understood the question about
"cloud-init" as "engine-init", once I answered yes to cloud-init, I was
allowed to set a root password and run engine-setup.

I am curious about some of your questions too, actually.

I was following the instructions from here:
http://www.ovirt.org/blog/2016/08/up-and-running-with-ovirt-4-0-and-gluster-storage/

In that, he said that in order to enable the gluster support in the engine,
one has to do it manually.  It seemed odd, or perhaps a misunderstanding of
the procedure.

My understanding was that after deploying my three hosts and gluster, I had
to deploy the hosted-appliance manually on host 1 so that after the initial
deploy, but before it was taken over by HA, I could go in and set the
gluster service to on, then finish the setup with the reboot, etc.

The other thing I was wondering about was setting up the additional hosts.
The instructions say to ssh to each host and deploy the engine there
through the ssh command line.  The tool itself says that I shouldn't be
doing it, I should be running through the website.  However, when I did
that, I had a number of issues that I wasn't able to correct, and when I
did it through ssh, I had the best functioning ovirt build yet.

One of the issues I had was when I tried to create a private network for
gluster sync'ing, the web interface saw the gluster IPs for hosts 2 and 3
(although host 1 was added with its ovirt/real management IP, and the two
networks were not routed).  The web itnerface host add ended up adding
hosts 2 and 3 with the gluster IP, and thus things broke a lot.  I had no
means of overriding those settings that I saw through the web interface.
When adding through SSH, I had a lot more control over adding the hosts,
and was able to add them "correctly".

I suspect this means the overall procedure was in error, and I'd like to
learn the "better" way to do it.

--Jim

On Mon, Jan 2, 2017 at 3:38 AM, Simone Tiraboschi 
wrote:

>
>
> On Mon, Jan 2, 2017 at 9:39 AM, Sandro Bonazzola 
> wrote:
>
>>
>>
>> On Fri, Dec 30, 2016 at 7:31 PM, Jim Kusznir  wrote:
>>
>>> Hi all:
>>>
>>> I'm trying to set up a new ovirt cluster.  I got it "mostly working"
>>> earlier, but wanted to change some physical networking stuff, and so I
>>> thought I'd blow away my machines and rebuild.  I followed the same recipe
>>> to build it all, but now I'm failing at a point that previously worked.
>>>
>>> I've built a 3 node cluster with glusterfs backing (3 brick replica),
>>> and all that is good and well.  I run the engine-setup --deploy, and it
>>> does its stuff, asks me (among other things) the admin password, I type in
>>> the password I want it to use (just like last time), then it says to log
>>> into the new VM and run engine-setup.  Here's the problem: I try to ssh in
>>> as root, and it will NOT accept my password.  It worked a couple days ago,
>>> doing it the exact same way, but it will not work now.
>>>
>>
> Hi Jim,
> sorry but why do you need to manually run engine-setup in the engine VM?
> If you are deploying with the ovirt-engine-appliance, hosted-engine-setup
> will run it for you with the right parameters.
>
>
>> I've destroyed and re-deployed several times, I've even done a low level
>>> wipe of all three nodes and rebuild everything, and again, it doesn't work.
>>>
>>> My only guess is that one of the packages the gdeploy script changed,
>>> and it has a bug or "new feature" that breaks this for some reason.
>>> Unfortunately, I do not have the package versions that worked or the
>>> current list to compare to, so I cannot support this.
>>>
>>> In any case, I'm completely stuck here...I can't log in to run
>>> engine-deploy, and I don't know enough of the console/low level stuff to
>>> try and hack my way into the VM (eg, to manually mount the disk image and
>>> replace the password or put my SSH key in).
>>>
>>> Suggestions?  Can anyone else replicate this?
>>>
>>
>> Can you please provide logs?
>>
>>
>>
>>>
>>> --Jim
>>>
>>> ___
>>> Users mailing list
>>> Users@ovirt.org
>>> http://lists.ovirt.org/mailman/listinfo/users
>>>
>>>
>>
>>
>> --
>> Sandro Bonazzola
>> 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: [ovirt-users] creating a vlan-tagged network

[Jim Kusznir]

Actually, I finally was able to identify the issue and fix it...Turns out
(as you probably expected), it wasn't ovirt...

My upstream provider had some wierd security left over, it limited the MAC
addresses permitted to exit the building, and my ovirt host made the list
somehow while my VMs did not.

I now have two VMs on two different nodes that are online!

Thank you for your help!

--Jim

On Sun, Jan 1, 2017 at 11:57 PM, Edward Haas  wrote:

>
>
> On Sun, Jan 1, 2017 at 7:16 PM, Jim Kusznir  wrote:
>
>> I pinged both the router on the subnet and a host IP in-between the two
>> ip's.
>>
>> [root@ovirt3 ~]# ping -I 162.248.147.33 162.248.147.1
>> PING 162.248.147.1 (162.248.147.1) from 162.248.147.33 : 56(84) bytes of
>> data.
>> 64 bytes from 162.248.147.1: icmp_seq=1 ttl=255 time=8.17 ms
>> 64 bytes from 162.248.147.1: icmp_seq=2 ttl=255 time=7.47 ms
>> 64 bytes from 162.248.147.1: icmp_seq=3 ttl=255 time=7.53 ms
>> 64 bytes from 162.248.147.1: icmp_seq=4 ttl=255 time=8.42 ms
>> ^C
>> --- 162.248.147.1 ping statistics ---
>> 4 packets transmitted, 4 received, 0% packet loss, time 3004ms
>> rtt min/avg/max/mdev = 7.475/7.901/8.424/0.420 ms
>> [root@ovirt3 ~]#
>>
>> The VM only has its public IP.
>>
>> --Jim
>>
>
> Very strange, all looks good to me.
>
> I can try to help you debug using tcpdump, just send me the details for
> remote connection on private.
> It will also help if you join the vdsm or ovir IRC channels.
>
>
>>
>> On Jan 1, 2017 01:26, "Edward Haas"  wrote:
>>
>>>
>>>
>>> On Sun, Jan 1, 2017 at 10:50 AM, Jim Kusznir 
>>> wrote:
>>>
 I currently only have two IPs assigned to me...I can try and take
 another, but that may not route out of the rack.  I've got the VM on one of
 the IPs and the host on the other currently.

 The switch is a "web-managed" basic 8-port switch (thrown in for
 testing while the "real" switch is in transit).  It has the 3 ports the
 hosts are plugged in configured with vlan 1 untagged, set as PVID, and vlan
 2 tagged.  Another port on the switch is untagged on vlan 1 connected to
 the router for the ovirtmgmt network (protected by a VPN, but not "burning"
 public IPs for mgmt purposes), another couple ports are untagged on vlan
 2.  One of those ports goes out of the rack, another goes to the router's
 internet port.  Router gets to the internet just fine.

 VM:
 kusznir@FusionPBX:~$ ip address
 1: lo:  mtu 65536 qdisc noqueue state UNKNOWN
 group default
 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
 inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
 inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
 2: eth0:  mtu 1500 qdisc pfifo_fast
 state UP group default qlen 1000
 link/ether 00:1a:4a:16:01:51 brd ff:ff:ff:ff:ff:ff
 inet 162.248.147.31/24 brd 162.248.147.255 scope global eth0
valid_lft forever preferred_lft forever
 inet6 fe80::21a:4aff:fe16:151/64 scope link
valid_lft forever preferred_lft forever
 kusznir@FusionPBX:~$ ip route
 default via 162.248.147.1 dev eth0
 162.248.147.0/24 dev eth0  proto kernel  scope link  src
 162.248.147.31
 kusznir@FusionPBX:~$

 Host:
 [root@ovirt3 ~]# ip address
 1: lo:  mtu 65536 qdisc noqueue state UNKNOWN
 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
 inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
 inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
 2: em1:  mtu 1500 qdisc mq master
 ovirtmgmt state UP qlen 1000
 link/ether 00:21:9b:98:2f:44 brd ff:ff:ff:ff:ff:ff
 3: em2:  mtu 1500 qdisc mq state DOWN qlen 1000
 link/ether 00:21:9b:98:2f:46 brd ff:ff:ff:ff:ff:ff
 4: em3:  mtu 1500 qdisc mq state DOWN qlen 1000
 link/ether 00:21:9b:98:2f:48 brd ff:ff:ff:ff:ff:ff
 5: em4:  mtu 1500 qdisc mq state
 DOWN qlen 1000
 link/ether 00:21:9b:98:2f:4a brd ff:ff:ff:ff:ff:ff
 6: ;vdsmdummy;:  mtu 1500 qdisc noop state DOWN
 link/ether 8e:1b:51:60:87:55 brd ff:ff:ff:ff:ff:ff
 7: ovirtmgmt:  mtu 1500 qdisc noqueue
 state UP
 link/ether 00:21:9b:98:2f:44 brd ff:ff:ff:ff:ff:ff
 inet 192.168.8.13/24 brd 192.168.8.255 scope global dynamic
 ovirtmgmt
valid_lft 54830sec preferred_lft 54830sec
 11: em1.2@em1:  mtu 1500 qdisc
 noqueue master Public_Cable state UP
 link/ether 00:21:9b:98:2f:44 brd ff:ff:ff:ff:ff:ff
 12: Public_Cable:  mtu 1500 qdisc
 noqueue state UP
 link/ether 00:21:9b:98:2f:44 brd ff:ff:ff:ff:ff:ff
 inet 162.248.147.33/24 brd 162.248.147.255 scope global
 Public_Cable
valid_lft forever preferred_lft forever
 14: vnet0:  mtu 1500 qdisc pfifo_fast
 master ovirtmgmt state UNKNOWN qlen 500
 link/ether fe:1a:4a:16:01:5

Re: [ovirt-users] Overlapping packages in CentOS 7 repo files from ovirt and mirror.centos.org

[Sandro Bonazzola]

On Mon, Jan 2, 2017 at 3:41 PM, Richard Chan 
wrote:

>
> "The fact you see overlapping versions in oVirt repo and CentOS
> Virtualization SIG repositories is due to the fact the two repositories are
> independent. You can use either the oVIrt one (built by oVirt release
> engineering) or the CentOS one (built by CentOS Virt SIG)."
>
>
> We are actually ok with using resources.ovirt.org packages but the CentOS
> VIrtualization SIG repo was enabled by ovirt-release40-4.0.5-2.noarch
> itself:
>
>
> /etc/yum.repos.d/ovirt-4.0-dependencies.repo:
>
> [centos-ovirt40-release]
> name=CentOS-7 - oVirt 4.0
> baseurl=http://mirror.centos.org/centos/7/virt/$basearch/ovirt-4.0/
> gpgcheck=0
> enabled=1
>
> This stanza does not use priority= or includepkgs= so is it deterministic
> that vdsm comes from resources.ovirt.org?
>
> 1. When I did a test run of "yum install vdsm":
>  vdsm-* packages were pulled from ovirt-4.0 (yay!)
> but
>   python-*
>   safelease
>   ioprocess
>   openvswitch
>   ovirt-imageio-*
> came from centos-ovirt40-release.
>
> 2. To be deterministic, i.e., force vdsm* from ovirt-4.0 insteaad of
> accidentally getting from CentOS SIG,  I am considering (similar to the
> epel repo stanza)
>
> [centos-ovirt40-release]
> name=CentOS-7 - oVirt 4.0
> baseurl=http://mirror.centos.org/centos/7/virt/$basearch/ovirt-4.0/
> includepkgs=python-* safelease ioprocess openvswitch ovirt-imageio-*
> ## or exclude=vdsm*
> gpgcheck=0
> enabled=1
>
>
> What do you think about this tightening? Thanks.
>

I think exclude vdsm* should be enough for your purpose.



>
>
>
>> Richard Chan
>
>


-- 
Sandro Bonazzola
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

Re: [ovirt-users] Overlapping packages in CentOS 7 repo files from ovirt and mirror.centos.org

[Richard Chan]

"The fact you see overlapping versions in oVirt repo and CentOS
Virtualization SIG repositories is due to the fact the two repositories are
independent. You can use either the oVIrt one (built by oVirt release
engineering) or the CentOS one (built by CentOS Virt SIG)."


We are actually ok with using resources.ovirt.org packages but the CentOS
VIrtualization SIG repo was enabled by ovirt-release40-4.0.5-2.noarch
itself:


/etc/yum.repos.d/ovirt-4.0-dependencies.repo:

[centos-ovirt40-release]
name=CentOS-7 - oVirt 4.0
baseurl=http://mirror.centos.org/centos/7/virt/$basearch/ovirt-4.0/
gpgcheck=0
enabled=1

This stanza does not use priority= or includepkgs= so is it deterministic
that vdsm comes from resources.ovirt.org?

1. When I did a test run of "yum install vdsm":
 vdsm-* packages were pulled from ovirt-4.0 (yay!)
but
  python-*
  safelease
  ioprocess
  openvswitch
  ovirt-imageio-*
came from centos-ovirt40-release.

2. To be deterministic, i.e., force vdsm* from ovirt-4.0 insteaad of
accidentally getting from CentOS SIG,  I am considering (similar to the
epel repo stanza)

[centos-ovirt40-release]
name=CentOS-7 - oVirt 4.0
baseurl=http://mirror.centos.org/centos/7/virt/$basearch/ovirt-4.0/
includepkgs=python-* safelease ioprocess openvswitch ovirt-imageio-*
## or exclude=vdsm*
gpgcheck=0
enabled=1


What do you think about this tightening? Thanks.



> Richard Chan
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users

Re: [ovirt-users] [HEADS UP] CentOS 7.3 is rolling out, need qemu-kvm-ev 2.6

[Michal Skrivanek]

> On 2 Jan 2017, at 13:52, Sandro Bonazzola  wrote:
> 
> 
> 
> On Mon, Jan 2, 2017 at 12:07 PM, Michal Skrivanek 
> mailto:michal.skriva...@redhat.com>> wrote:
> 
>> On 2 Jan 2017, at 12:01, Sandro Bonazzola > > wrote:
>> 
>> 
>> 
>> On Wed, Dec 28, 2016 at 9:55 AM, Gianluca Cecchi > > wrote:
>> 
>> As I wrote, no qemu package proposed. Because oVirt provided qemu is greater 
>> in version in respect with 7.3
>> In fact, as I already wrote, I have now 
>> qemu-kvm-common-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-img-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-kvm-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-kvm-tools-ev-2.3.0-31.el7.16.1.x86_64
>> 
>> Hi Gianluca, I added 2.6.0 to oVirt 4.0 repository at 
>> http://resources.ovirt.org/pub/ovirt-4.0/rpm/el7/ 
>> .
>> It was supposed to be published there with 4.0.6 GA coming ina few days but 
>> it has been already tested enough to be promoted to release repository.
> 
> it would be best to remove it. We really do not want to get out of sync with 
> platform
> 
> Can you please tell me more about this? 

we do not want to manually add platform-provided stuff in our repos. We should 
only put ovirt packages there, like ovirt-*, ioprocess and such. But everything 
else should be pulled in by external repos, even when it is us who maintain 
them. Otherwise we often run into conflicts e.g. when platform releases an 
update of a dependency and we do not release it at the same time in our repo.

> 
>  
> It is already causing problems since qemu-kvm-rhev package is 2.6.0-27 and 
> the one you placed in resources.ovirt.org  is 
> 2.6.0-27.1 (sure, it’s not a supported combination, but it complicates 
> development)
> 
> I'm not sure what's the issue you're describing here.
> qemu-kvm-ev is 2.6.0-27.1 being 2.6.0-27 +2 patches (rebranding and dep 
> fixing) so it's not the same release. Also qemu-kvm-ev is supposed to provide 
> qemu-kvm-rhev.

yes, and when you have qemu-kvm-rhev with ovirt repos it will replace all -rhev 
packages with -ev because it thinks it’s newer. Not a big deal as indeed it is 
the same, but it confuses people…

Thanks,
michal

> 
> 
> 
>  
> 
>> 
>> For having a qemu-kvm-ev version always aligned with CentOS release I would 
>> suggest to run:
>> "yum install centos-release-qemu-ev" on your CentOS hosts.
>> 
>> If you decide to use CentOS VIrt SIG provided packages on your hosts 
>> (excluding the oVirt engine host which still require oVirt repositories) I 
>> would suggest to remove ovirt-release rpms and /etc/yum.repos.d/ovirt-*.repo 
>> and then run:
>> "yum install centos-release-ovirt40".
>> 
>> this will add as dependencies:
>> Requires: centos-release-virt-common
>> Requires: centos-release-qemu-ev
>> Requires: centos-release-gluster37
>> 
>> providing the latest gluster 3.7 and latest qemu-kvm-ev for your hosts.
>> 
>> Having both oVirt repos and Virt SIG repos enabled should still work being 
>> both the repos built from the same sources but it's not recommended.
>> 
>> 
>> -- 
>> Sandro Bonazzola
>> 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 
>> 
> 
> 
> 
> 
> -- 
> Sandro Bonazzola
> 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

Re: [ovirt-users] [HEADS UP] CentOS 7.3 is rolling out, need qemu-kvm-ev 2.6

[Sandro Bonazzola]

On Mon, Jan 2, 2017 at 12:07 PM, Michal Skrivanek <
michal.skriva...@redhat.com> wrote:

>
> On 2 Jan 2017, at 12:01, Sandro Bonazzola  wrote:
>
>
>
> On Wed, Dec 28, 2016 at 9:55 AM, Gianluca Cecchi <
> gianluca.cec...@gmail.com> wrote:
>
>>
>> As I wrote, no qemu package proposed. Because oVirt provided qemu is
>> greater in version in respect with 7.3
>> In fact, as I already wrote, I have now
>> qemu-kvm-common-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-img-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-kvm-ev-2.3.0-31.el7.16.1.x86_64
>> qemu-kvm-tools-ev-2.3.0-31.el7.16.1.x86_64
>>
>
> Hi Gianluca, I added 2.6.0 to oVirt 4.0 repository at
> http://resources.ovirt.org/pub/ovirt-4.0/rpm/el7/.
> It was supposed to be published there with 4.0.6 GA coming ina few days
> but it has been already tested enough to be promoted to release repository.
>
>
> it would be best to remove it. We really do not want to get out of sync
> with platform
>

Can you please tell me more about this?



> It is already causing problems since qemu-kvm-rhev package is 2.6.0-27 and
> the one you placed in resources.ovirt.org is 2.6.0-27.1 (sure, it’s not a
> supported combination, but it complicates development)
>

I'm not sure what's the issue you're describing here.
qemu-kvm-ev is 2.6.0-27.1 being 2.6.0-27 +2 patches (rebranding and dep
fixing) so it's not the same release. Also qemu-kvm-ev is supposed to
provide qemu-kvm-rhev.





>
>
> For having a qemu-kvm-ev version always aligned with CentOS release I
> would suggest to run:
> "yum install centos-release-qemu-ev" on your CentOS hosts.
>
> If you decide to use CentOS VIrt SIG provided packages on your hosts
> (excluding the oVirt engine host which still require oVirt repositories) I
> would suggest to remove ovirt-release rpms and /etc/yum.repos.d/ovirt-*.repo
> and then run:
> "yum install centos-release-ovirt40".
>
> this will add as dependencies:
> Requires: centos-release-virt-common
> Requires: centos-release-qemu-ev
> Requires: centos-release-gluster37
>
> providing the latest gluster 3.7 and latest qemu-kvm-ev for your hosts.
>
> Having both oVirt repos and Virt SIG repos enabled should still work being
> both the repos built from the same sources but it's not recommended.
>
>
> --
> Sandro Bonazzola
> 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
>
>
>


-- 
Sandro Bonazzola
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

Re: [ovirt-users] Issue with OVN/OVS and mandatory ovirtmgmt network

[Sverker Abrahamsson]

Got it to work now by following the env8 example in OVN tutorial, where 
a port is added with type l2gateway. Not sure how that is different from 
the localnet variant, but didn't suceed in getting that one working. Now 
I'm able to ping and telnet over the tunnel, but not ssh even when the 
port is answering on telnet. Neither does nfs traffic work even though 
mount did. Suspecting MTU issue. I did notice that ovn-controller starts 
too early, before network interfaces are established and hence can't 
reach the db. As these is a purely OVS/OVN issue I'll ask about it on 
their mailing list.


Getting back to the original issue with Ovirt, I've now added the second 
host h1 to ovirt-engine. Had to do the same as with h2 to create a dummy 
ovirtmgmt network but configured access via the public IP. My firewall 
settings was replaced with iptables config and vdsm.conf was overwritten 
when engine was set up, so those had to be manually restored. It would 
be preferable if it would be possible to configure ovirt-engine that it 
does not "own" the host and instead comply with the settings it has 
instead of enforcing it's own view..


Apart from that it seems the second host works, although I need to 
resolve the traffic issue over the OVS tunnel.

/Sverker

Den 2017-01-02 kl. 01:13, skrev Sverker Abrahamsson:
1. That is not possible as ovirt (or vdsm) will rewrite the network 
configuration to a non-working state. That is why I've set that if as 
hidden to vdsm and is why I'm keen on getting OVS/OVN to work


2. I've been reading the doc for OVN and starting to connect the dots, 
which is not trivial as it is complex. Some insights reached:


First step is the OVN database, installed by openvswitch-ovn-central, 
which I currently have running on h2 host. The 'ovn-nbctl' and 
'ovn-sbctl' commands are only possible to execute on a database node. 
Two ip's are given to 'vdsm-tool ovn-config  ip>' as arguments, where  is how this OVN node reaches 
the database and  is the ip to which other OVN nodes sets 
up a tunnel to this node. I.e. it is not for creating a tunnel to the 
database which I thought first from the description in blog post.


The tunnel between OVN nodes is of type geneve which is a UDP based 
protocol but I have not been able to find anywhere which port is used 
so that I can open it in firewalld. I have added OVN on another host, 
called h1, and connected it to the db. I see there is traffic to the 
db port, but I don't see any geneve traffic between the nodes.


Ovirt is now able to create it's vnet0 interface on the br-int ovs 
bridge, but then I run into the next issue. How do I create a 
connection from the logical switch to the physical host? I need that 
to a) get a connection out to the internet through a masqueraded if or 
ipv6 and b) be able to run a dhcp server to give ip's to the VM's.


/Sverker

Den 2016-12-30 kl. 18:05, skrev Marcin Mirecki:

1. Why not use your physical nic for ovirtmgmt then?

2. "ovn-nbctl ls-add" does not add a bridge, but a logical switch.
br-int is an internal OVN implementation detail, which the user
should not care about. What you see in the ovirt UI are logical
networks. They are implemented as OVN logical switches in case
of the OVN provider.

Please look at:
http://www.ovirt.org/blog/2016/11/ovirt-provider-ovn/
You can get the latest rpms from here:
http://resources.ovirt.org/repos/ovirt/experimental/master/ovirt-provider-ovn_fc24_46/rpm/fc24/noarch/ 



- Original Message -

From: "Sverker Abrahamsson" 
To: "Marcin Mirecki" 
Cc: "Ovirt Users" 
Sent: Friday, December 30, 2016 4:25:58 PM
Subject: Re: [ovirt-users] Issue with OVN/OVS and mandatory 
ovirtmgmt network


1. No, I did not want to put the ovirtmgmt bridge on my physical nic as
it always messed up the network config making the host unreachable. I
have put a ovs bridge on this nic which I will use to make tunnels when
I add other hosts. Maybe br-int will be used for that instead, will see
when I get that far.

As it is now I have a dummy if for ovirtmgmt bridge but this will
probably not work when I add other hosts as that bridge cannot connect
to the other hosts. I'm considering keeping this just as a dummy to 
keep

ovirt engine satisfied while the actual communication will happen over
OVN/OVS bridges and tunnels.

2. On
https://www.ovirt.org//develop/release-management/features/ovirt-ovn-provider/ 


there is instructions how to add an OVS bridge to OVN with |ovn-nbctl
ls-add |. If you want to use br-int then it makes 
sense to
make that bridge visible in ovirt webui under networks so that it 
can be

selected for VM's.

It quite doesn't make sense to me that I can select other network 
for my

VM but then that setting is not used when setting up the network.

/Sverker

Den 2016-12-30 kl. 15:34, skrev Marcin Mirecki:

Hi,

The OVN provider does not require you to add any bridges manually.
As I understand we were dealing with two problems:
1. You only had one physical ni

Re: [ovirt-users] Using zRam with oVirt Nodes

[Fernando Frediani]

All nodes I user zSwap (or Swap in zRam) I do have also some swap space 
in disk with lower priority which when/if gets used means something warning.


Overall the usage of zSwap helps a lot to increase the density of a Host 
at the cost of CPU which in most cases is more idle than memory or storage


Fernando


On 30/12/2016 21:47, Pavel Gashev wrote:

I enable zSwap on my oVirt nodes by default. However zSwap is not the same 
thing as zRam. The purpose of zSwap is decreasing I/O by compressing swap 
itself. zSwap uses some ram for compressed swap, but it doesn’t try to keep a 
page in memory if it can’t be compressed.

I find using of zRam without swap too risky. You don’t know when OOM killer 
start its job.

-Original Message-
From:  on behalf of Fernando Frediani 

Date: Friday 30 December 2016 at 20:06
To: "users@ovirt.org" 
Subject: [ovirt-users] Using zRam with oVirt Nodes

Hello folks.

On simple libvirt/KVM hosts in order to improve RAM usage and avoid swap
to disk I use zRam with swap to RAM technique. So create half of amount
of host memory in zRam disk divided by the number of CPU cores. Works
pretty well.

Has anyone tried it with oVirt Nodes or even has it been considered as a
feature to simply turn on/off ?

Thanks
Fernando

___
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: [ovirt-users] New install: can't install engine

[Simone Tiraboschi]

On Mon, Jan 2, 2017 at 9:39 AM, Sandro Bonazzola 
wrote:

>
>
> On Fri, Dec 30, 2016 at 7:31 PM, Jim Kusznir  wrote:
>
>> Hi all:
>>
>> I'm trying to set up a new ovirt cluster.  I got it "mostly working"
>> earlier, but wanted to change some physical networking stuff, and so I
>> thought I'd blow away my machines and rebuild.  I followed the same recipe
>> to build it all, but now I'm failing at a point that previously worked.
>>
>> I've built a 3 node cluster with glusterfs backing (3 brick replica), and
>> all that is good and well.  I run the engine-setup --deploy, and it does
>> its stuff, asks me (among other things) the admin password, I type in the
>> password I want it to use (just like last time), then it says to log into
>> the new VM and run engine-setup.  Here's the problem: I try to ssh in as
>> root, and it will NOT accept my password.  It worked a couple days ago,
>> doing it the exact same way, but it will not work now.
>>
>
Hi Jim,
sorry but why do you need to manually run engine-setup in the engine VM?
If you are deploying with the ovirt-engine-appliance, hosted-engine-setup
will run it for you with the right parameters.


> I've destroyed and re-deployed several times, I've even done a low level
>> wipe of all three nodes and rebuild everything, and again, it doesn't work.
>>
>> My only guess is that one of the packages the gdeploy script changed, and
>> it has a bug or "new feature" that breaks this for some reason.
>> Unfortunately, I do not have the package versions that worked or the
>> current list to compare to, so I cannot support this.
>>
>> In any case, I'm completely stuck here...I can't log in to run
>> engine-deploy, and I don't know enough of the console/low level stuff to
>> try and hack my way into the VM (eg, to manually mount the disk image and
>> replace the password or put my SSH key in).
>>
>> Suggestions?  Can anyone else replicate this?
>>
>
> Can you please provide logs?
>
>
>
>>
>> --Jim
>>
>> ___
>> Users mailing list
>> Users@ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
>>
>>
>
>
> --
> Sandro Bonazzola
> 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: [ovirt-users] [HEADS UP] CentOS 7.3 is rolling out, need qemu-kvm-ev 2.6

[Michal Skrivanek]

> On 2 Jan 2017, at 12:01, Sandro Bonazzola  wrote:
> 
> 
> 
> On Wed, Dec 28, 2016 at 9:55 AM, Gianluca Cecchi  > wrote:
> 
> As I wrote, no qemu package proposed. Because oVirt provided qemu is greater 
> in version in respect with 7.3
> In fact, as I already wrote, I have now 
> qemu-kvm-common-ev-2.3.0-31.el7.16.1.x86_64
> qemu-img-ev-2.3.0-31.el7.16.1.x86_64
> qemu-kvm-ev-2.3.0-31.el7.16.1.x86_64
> qemu-kvm-tools-ev-2.3.0-31.el7.16.1.x86_64
> 
> Hi Gianluca, I added 2.6.0 to oVirt 4.0 repository at 
> http://resources.ovirt.org/pub/ovirt-4.0/rpm/el7/ 
> .
> It was supposed to be published there with 4.0.6 GA coming ina few days but 
> it has been already tested enough to be promoted to release repository.

it would be best to remove it. We really do not want to get out of sync with 
platform
It is already causing problems since qemu-kvm-rhev package is 2.6.0-27 and the 
one you placed in resources.ovirt.org is 2.6.0-27.1 (sure, it’s not a supported 
combination, but it complicates development)

> 
> For having a qemu-kvm-ev version always aligned with CentOS release I would 
> suggest to run:
> "yum install centos-release-qemu-ev" on your CentOS hosts.
> 
> If you decide to use CentOS VIrt SIG provided packages on your hosts 
> (excluding the oVirt engine host which still require oVirt repositories) I 
> would suggest to remove ovirt-release rpms and /etc/yum.repos.d/ovirt-*.repo 
> and then run:
> "yum install centos-release-ovirt40".
> 
> this will add as dependencies:
> Requires: centos-release-virt-common
> Requires: centos-release-qemu-ev
> Requires: centos-release-gluster37
> 
> providing the latest gluster 3.7 and latest qemu-kvm-ev for your hosts.
> 
> Having both oVirt repos and Virt SIG repos enabled should still work being 
> both the repos built from the same sources but it's not recommended.
> 
> 
> -- 
> Sandro Bonazzola
> 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: [ovirt-users] [HEADS UP] CentOS 7.3 is rolling out, need qemu-kvm-ev 2.6

[Sandro Bonazzola]

On Wed, Dec 28, 2016 at 9:55 AM, Gianluca Cecchi 
wrote:

>
> As I wrote, no qemu package proposed. Because oVirt provided qemu is
> greater in version in respect with 7.3
> In fact, as I already wrote, I have now
> qemu-kvm-common-ev-2.3.0-31.el7.16.1.x86_64
> qemu-img-ev-2.3.0-31.el7.16.1.x86_64
> qemu-kvm-ev-2.3.0-31.el7.16.1.x86_64
> qemu-kvm-tools-ev-2.3.0-31.el7.16.1.x86_64
>

Hi Gianluca, I added 2.6.0 to oVirt 4.0 repository at
http://resources.ovirt.org/pub/ovirt-4.0/rpm/el7/.
It was supposed to be published there with 4.0.6 GA coming ina few days but
it has been already tested enough to be promoted to release repository.

For having a qemu-kvm-ev version always aligned with CentOS release I would
suggest to run:
"yum install centos-release-qemu-ev" on your CentOS hosts.

If you decide to use CentOS VIrt SIG provided packages on your hosts
(excluding the oVirt engine host which still require oVirt repositories) I
would suggest to remove ovirt-release rpms and /etc/yum.repos.d/ovirt-*.repo
and then run:
"yum install centos-release-ovirt40".

this will add as dependencies:
Requires: centos-release-virt-common
Requires: centos-release-qemu-ev
Requires: centos-release-gluster37

providing the latest gluster 3.7 and latest qemu-kvm-ev for your hosts.

Having both oVirt repos and Virt SIG repos enabled should still work being
both the repos built from the same sources but it's not recommended.


-- 
Sandro Bonazzola
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

Re: [ovirt-users] HA Score

[knarra]

On 01/02/2017 02:59 PM, Simone Tiraboschi wrote:



On Mon, Jan 2, 2017 at 9:23 AM, knarra > wrote:


On 01/01/2017 06:23 PM, Doron Fediuck wrote:

Care to share the HE agents log files?

Hi Doron,

I have collected  logs from all the machines and placed it in
the link below.

http://rhsqe-repo.lab.eng.blr.redhat.com/sosreports/HC/ovirt_logs/



From the logs it seams that host 1 and host 3 are stable at 3400 
points while host 2 periodically goes down to 1800 due to:


MainThread::INFO::2017-01-01 
17:26:33,324::states::128::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(score) 
Penalizing score by 1600 due to gateway status
Thanks simone. So i understand that problem is because of bad gateway in 
Host2.




Thanks
kasturi

On Fri, Dec 30, 2016 at 9:17 AM, knarra mailto:kna...@redhat.com>> wrote:

Hi,

 I have latest 4.1 installed and i see that HA score
on hosts keeps going
to 0  and comes back to  3400. This behavior is something
which i am
observing with 4.1  and i see that it takes considerable
amount of time to
get back to normal state. Any reason why it takes such a
long time?

Thanks

kasturi.

___
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: [ovirt-users] Fwd: [Call for feedback] anybody gave 4.1 beta a try?

[Martin Polednik]

On 23/12/16 10:42 +0100, Nathanaël Blanchet wrote:



Le 22/12/2016 à 17:17, Yaniv Kaul a écrit :



On Thu, Dec 22, 2016 at 12:50 PM, Nathanaël Blanchet 
mailto:blanc...@abes.fr>> wrote:


   Hi,

   sebool unconfigured module prevents vdsm from starting:


We always run with selinux. What's the reason for disabling it?
I never disabled selinux, why do you think I did this? it is active, 
this is the problem!

Y.


   [root@gaua2 network-scripts]# vdsm-tool configure --force

   Checking configuration status...

   Current revision of multipath.conf detected, preserving
   libvirt is already configured for vdsm
   SUCCESS: ssl configured to true. No conflicts

   Running configure...
   libsepol.context_from_record: type ovirt_vmconsole_host_port_t is
   not defined (No such file or directory).
   libsepol.context_from_record: could not create context structure
   (Invalid argument).
   libsepol.port_from_record: could not create port structure for
   range 2223:2223 (tcp) (Invalid argument).
   libsepol.sepol_port_modify: could not load port range 2223 - 2223
   (tcp) (Invalid argument).
   libsemanage.dbase_policydb_modify: could not modify record value
   (Invalid argument).
   libsemanage.semanage_base_merge_components: could not merge local
   modifications into policy (Invalid argument).
   Traceback (most recent call last):
 File "/usr/bin/vdsm-tool", line 219, in main
   return tool_command[cmd]["command"](*args)
 File "/usr/lib/python2.7/site-packages/vdsm/tool/__init__.py",
   line 38, in wrapper
   func(*args, **kwargs)
 File
   "/usr/lib/python2.7/site-packages/vdsm/tool/configurator.py", line
   141, in configure
   _configure(c)
 File
   "/usr/lib/python2.7/site-packages/vdsm/tool/configurator.py", line
   88, in _configure
   getattr(module, 'configure', lambda: None)()
 File
   "/usr/lib/python2.7/site-packages/vdsm/tool/configurators/sebool.py",
   line 84, in configure
   _setup_booleans(True)
 File
   "/usr/lib/python2.7/site-packages/vdsm/tool/configurators/sebool.py",
   line 57, in _setup_booleans
   sebool_obj.finish()
 File "/usr/lib/python2.7/site-packages/seobject/__init__.py",
   line 320, in finish
   self.commit()
 File "/usr/lib/python2.7/site-packages/seobject/__init__.py",
   line 310, in commit
   rc = semanage_commit(self.sh)
   OSError: [Errno 22] Invalid argument


Had same issue, managed to solve it by reinstalling ovirt_vmconsole
rules:

# semodule -i /usr/share/selinux/packages/ovirt-vmconsole/ovirt_vmconsole.pp

mpolednik



   Le 19/12/2016 à 11:55, Maton, Brett a écrit :

   I'll give that ago this evening, although that might be related
   to trying to deploy to the host multiple times

   On 19 December 2016 at 09:54, Simone Tiraboschi
   mailto:stira...@redhat.com>> wrote:



   On Mon, Dec 19, 2016 at 10:11 AM, Maton, Brett
   mailto:mat...@ltresources.co.uk>>
   wrote:

   Log from the last deploy attempt attached.


   2016-12-18 12:52:05 DEBUG
   otopi.plugins.ovirt_host_mgmt.packages.update
   update.verbose:94 Yum: There are unfinished transactions
   remaining. You might consider running
   yum-complete-transaction, or "yum-complete-transaction
   --cleanup-only" and "yum history redo last", first to finish
   them. If those don't work you'll have to try
   removing/installing packages by hand (maybe package-cleanup
   can help).

   It seams that you have unfinished yum transactions on your
   host; can you please check for that and then try redeploying
   the host?


   On 19 December 2016 at 08:30, Simone Tiraboschi
   mailto:stira...@redhat.com>> wrote:



   On Sat, Dec 17, 2016 at 7:17 PM, Maton, Brett
   mailto:mat...@ltresources.co.uk>> wrote:

   I had a mac address conflict so reinstalled (
   from scratch ) 4.1 this morning, CPU load issue
   in UI was still there.


   We already have an open bug about this:
   https://bugzilla.redhat.com/show_bug.cgi?id=1390675
   Thanks for the report.


   I can't deploy additional hosts now though due to
   SSH errors.


   Could you please attach host-deploy logs?


   Experimenting, I updated the hosted engine this
   afternoon and the CPU load issue seems be have
   either been fixed or resolved itself.

   However, SSH errors deploying additional hosts
   persists ( I had no problem with the previous
   install on Friday ).

   enging.log snippet:
   2016-12-17 17:09:54,764Z ERROR
   [org.ovirt.engine.core.uutils.ssh.SSHDialog]
   (org.ovirt.thread.pool-6-thread-26)
   [b04a9eaf-65b5-4a80-a254-0b29a8fea573] SSH error

Re: [ovirt-users] HA Score

[Simone Tiraboschi]

On Mon, Jan 2, 2017 at 9:23 AM, knarra  wrote:

> On 01/01/2017 06:23 PM, Doron Fediuck wrote:
>
>> Care to share the HE agents log files?
>>
> Hi Doron,
>
> I have collected  logs from all the machines and placed it in the link
> below.
>
> http://rhsqe-repo.lab.eng.blr.redhat.com/sosreports/HC/ovirt_logs/


>From the logs it seams that host 1 and host 3 are stable at 3400 points
while host 2 periodically goes down to 1800 due to:

MainThread::INFO::2017-01-01
17:26:33,324::states::128::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(score)
Penalizing score by 1600 due to gateway status


>
>
> Thanks
> kasturi
>
>> On Fri, Dec 30, 2016 at 9:17 AM, knarra  wrote:
>>
>>> Hi,
>>>
>>>  I have latest 4.1 installed and i see that HA score on hosts keeps
>>> going
>>> to 0  and comes back to  3400. This behavior is something which i am
>>> observing with 4.1  and i see that it takes considerable amount of time
>>> to
>>> get back to normal state. Any reason why it takes such a long time?
>>>
>>> Thanks
>>>
>>> kasturi.
>>>
>>> ___
>>> 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: [ovirt-users] New install: can't install engine

[Sandro Bonazzola]

On Fri, Dec 30, 2016 at 7:31 PM, Jim Kusznir  wrote:

> Hi all:
>
> I'm trying to set up a new ovirt cluster.  I got it "mostly working"
> earlier, but wanted to change some physical networking stuff, and so I
> thought I'd blow away my machines and rebuild.  I followed the same recipe
> to build it all, but now I'm failing at a point that previously worked.
>
> I've built a 3 node cluster with glusterfs backing (3 brick replica), and
> all that is good and well.  I run the engine-setup --deploy, and it does
> its stuff, asks me (among other things) the admin password, I type in the
> password I want it to use (just like last time), then it says to log into
> the new VM and run engine-setup.  Here's the problem: I try to ssh in as
> root, and it will NOT accept my password.  It worked a couple days ago,
> doing it the exact same way, but it will not work now.
>
> I've destroyed and re-deployed several times, I've even done a low level
> wipe of all three nodes and rebuild everything, and again, it doesn't work.
>
> My only guess is that one of the packages the gdeploy script changed, and
> it has a bug or "new feature" that breaks this for some reason.
> Unfortunately, I do not have the package versions that worked or the
> current list to compare to, so I cannot support this.
>
> In any case, I'm completely stuck here...I can't log in to run
> engine-deploy, and I don't know enough of the console/low level stuff to
> try and hack my way into the VM (eg, to manually mount the disk image and
> replace the password or put my SSH key in).
>
> Suggestions?  Can anyone else replicate this?
>

Can you please provide logs?



>
> --Jim
>
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
>


-- 
Sandro Bonazzola
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

Re: [ovirt-users] New install: can't install engine

[Sahina Bose]

On Sat, Dec 31, 2016 at 12:01 AM, Jim Kusznir  wrote:

> Hi all:
>
> I'm trying to set up a new ovirt cluster.  I got it "mostly working"
> earlier, but wanted to change some physical networking stuff, and so I
> thought I'd blow away my machines and rebuild.  I followed the same recipe
> to build it all, but now I'm failing at a point that previously worked.
>
> I've built a 3 node cluster with glusterfs backing (3 brick replica), and
> all that is good and well.  I run the engine-setup --deploy, and it does
> its stuff, asks me (among other things) the admin password, I type in the
> password I want it to use (just like last time), then it says to log into
> the new VM and run engine-setup.  Here's the problem: I try to ssh in as
> root, and it will NOT accept my password.  It worked a couple days ago,
> doing it the exact same way, but it will not work now.
>
> I've destroyed and re-deployed several times, I've even done a low level
> wipe of all three nodes and rebuild everything, and again, it doesn't work.
>
> My only guess is that one of the packages the gdeploy script changed, and
> it has a bug or "new feature" that breaks this for some reason.
> Unfortunately, I do not have the package versions that worked or the
> current list to compare to, so I cannot support this.
>

I doubt that gdeploy changes caused it. It primarily sets up the gluster
cluster and environment to invoke hosted-engine deployment.

Could be a change in the ovirt-engine appliance. Are you able to access the
VM console via vnc? Is the root login disabled in appliance?


>
> In any case, I'm completely stuck here...I can't log in to run
> engine-deploy, and I don't know enough of the console/low level stuff to
> try and hack my way into the VM (eg, to manually mount the disk image and
> replace the password or put my SSH key in).
>
> Suggestions?  Can anyone else replicate this?
>
> --Jim
>
> ___
> 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: [ovirt-users] HA Score

[knarra]

On 01/01/2017 06:23 PM, Doron Fediuck wrote:

Care to share the HE agents log files?

Hi Doron,

I have collected  logs from all the machines and placed it in the 
link below.


http://rhsqe-repo.lab.eng.blr.redhat.com/sosreports/HC/ovirt_logs/

Thanks
kasturi

On Fri, Dec 30, 2016 at 9:17 AM, knarra  wrote:

Hi,

 I have latest 4.1 installed and i see that HA score on hosts keeps going
to 0  and comes back to  3400. This behavior is something which i am
observing with 4.1  and i see that it takes considerable amount of time to
get back to normal state. Any reason why it takes such a long time?

Thanks

kasturi.

___
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