[ovirt-users] ovirt 4.1 selfhosted engine deployment

[Lauri-Alo Adamson]

Hello!

Thanks  - try to remember put cc 

More information:
Our server has separated management Network interfaces (nic) - in HP server i use ILO web managem interface to access server remote console. (HP iLO Integrated Remote Console)
I used this renote console to access this freezed node and i was no abel to  ping  or access ohter host in the same network. Then  open this remote console i saw many  "vcpu unhandled rdmsr" messages.
My node had GbE nic interface with ip address that was not used by ovirtmagmt interface. I was no tabel ping or access this interfiace then node freezed.

Lauri-Alo Adamson




-----Original Message-----
From: Yedidyah Bar David [mailto:didi at redhat.com] 
Sent: Tuesday, March 14, 2017 3:36 PM
To: Lauri-Alo Adamson <Lauri.Adamson at andmevara.ee>
Subject: Re: [ovirt-users] ovirt 4.1 selfhosted engine deployment

Hi,

Please keep the list in Cc, so that others can try to help, and/or learn. Thanks.

On Tue, Mar 14, 2017 at 3:14 PM, Lauri-Alo Adamson <Lauri.Adamson at andmevara.ee> wrote:
> Hello !
>
> Additional clarifications:
>
> Host/node freezed after the hosted engine virtuaal machine tried first 
> time to start. After that i was no tabel to Access node over ssh Whole node freezed. Only after reboot i was abel to access thise host again  over ssh.

You said you saw something on the console. I thought you meant you managed to login.
If you didn't, then indeed the problem wasn't (probably) only in networking, but something deeper, in the kernel probably.

Best,

>
> Lauri-Alo Adamson
>
> -----Original Message-----
> From: Yedidyah Bar David [mailto:didi at redhat.com]
> Sent: Tuesday, March 14, 2017 3:04 PM
> To: Lauri-Alo Adamson <Lauri.Adamson at andmevara.ee>
> Cc: users at ovirt.org
> Subject: Re: [ovirt-users] ovirt 4.1 selfhosted engine deployment
>
> On Tue, Mar 14, 2017 at 2:48 PM, Lauri-Alo Adamson <Lauri.Adamson at andmevara.ee> wrote:
>> Hello!
>>
>>
>>
>> I tried to install/deploy ovirt 4.1 self hosted engine  with 
>> glusterfs (thee node cluster two HP 380D Gen9 server with Intel(R) 
>> Xeon(R) CPU
>> E5-2650 v3 @ 2.30GHz and one IBM 3650 M4 ) on Centos 7.3.1611 x86-64 .
>> But then the hosted engine
>>
>> virtuaal machine is starting  then whole  host (node)  freezes – 
>> server is not accessible over network with ssh. Engine virtuaal 
>> machines fails to start and deploiment fails. Same time the on server 
>> console and server log apper lines „vcpu  unhandled rdmsr“ . I 
>> searched Google this problem and dis not found the solution.
>>
>>
>>
>> There was talk about  adding  /etc/modprobe.d/kvm.conf  „options kvm 
>> ignore_msrs=1“  Which is the same as „echo 1 > 
>> /sys/module/kvm/parameters/ignore_msrs“
>>
>> http://wiki.stoney-cloud.org/wiki/Workaround_unhandled_rdmsr/wrmsr
>>
>>
>>
>> I tried but these . Sadly they  do not work. Has any one found 
>> solution to the described probleem.
>
> Are you sure your problem isn't a simple networking problem? At what stage does it die? Do you run it under 'screen'? If so, you can reattach to it ('screen -r -d') and see where you are. Also please check/post the setup logs (/var/log/ovirt-hosted-engine-setup/*).
>
> If indeed your problem is kernel/kvm related, then I have no idea.
> Does starting a kvm vm using virsh/virt-install work?
>
> Good luck,
>
>>
>>
>>
>> Lauri-Alo Adamson
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> _______________________________________________
>> Users mailing list
>> Users at ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
>>
>
>
>
> --
> Didi



--
Didi