[ovirt-users] hosted-engine vm-status stale data and cluster seems "broken"

Tue Jun 14 12:16:27 UTC 2016

Martin -

One thing I noticed on all of the nodes is this:

Jun 14 08:11:11 njsevcnp01 ovirt-ha-agent[15713]: ovirt-ha-agent
ovirt_hosted_engine_ha.lib.brokerlink.BrokerLink ERROR Connection closed:
Connection timed out
Jun 14 08:11:11 njsevcnp01.fixflyer.com ovirt-ha-agent[15713]:
ovirt-ha-agent ovirt_hosted_engine_ha.agent.agent.Agent ERROR Error:
'Failed to start monitor <type 'type'>, options {'hostname': 'njsevcnp01'}:
Connection timed out' - trying to restart agent

Then the agent is restarted

[root at njsevcnp01 ~]# ps -Aef | grep -i ovirt-ha-agent | grep -iv grep
vdsm      15713      1  0 08:09 ?        00:00:01 /usr/bin/python
/usr/share/ovirt-hosted-engine-ha/ovirt-ha-agent --no-daemon

I dont know why the connection would time out because as you can see that
log is from node01 and I cant figure out why its timing out on the
connection

The other interesting thing is this log from node01. The odd thing is that
it seems there is some split brain somewhere in oVirt because this log is
from node02 but it is asking the engine and its getting back "vm not
running on this host' rather than 'stale data'. But I dont know engine
internals

MainThread::INFO::2016-06-14
08:13:05,163::state_machine::171::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(refresh)
Host njsevcnp02 (id 2): {hostname: njsevcnp02, host-id: 2, engine-status:
{reason: vm not running on this host, health: bad, vm: down, detail:
unknown}, score: 0, stopped: True, maintenance: False, crc32: 25da07df,
host-ts: 3030}
MainThread::INFO::2016-06-14
08:13:05,163::state_machine::171::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(refresh)
Host njsevcnp03 (id 3): {hostname: njsevcnp03, host-id: 3, engine-status:
{reason: vm not running on this host, health: bad, vm: down, detail:
unknown}, score: 0, stopped: True, maintenance: False, crc32: c67818cb,
host-ts: 10877406}

And that same log on node02 where the engine is running

MainThread::INFO::2016-06-14
08:15:44,451::state_machine::171::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(refresh)
Host njsevcnp01 (id 1): {hostname: njsevcnp01, host-id: 1, engine-status:
{reason: vm not running on this host, health: bad, vm: down, detail:
unknown}, score: 0, stopped: True, maintenance: False, crc32: 260dbf06,
host-ts: 327}
MainThread::INFO::2016-06-14
08:15:44,451::state_machine::171::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(refresh)
Host njsevcnp03 (id 3): {hostname: njsevcnp03, host-id: 3, engine-status:
{reason: vm not running on this host, health: bad, vm: down, detail:
unknown}, score: 0, stopped: True, maintenance: False, crc32: c67818cb,
host-ts: 10877406}
MainThread::INFO::2016-06-14
08:15:44,451::state_machine::174::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(refresh)
Local (id 2): {engine-health: {health: good, vm: up, detail: up}, bridge:
True, mem-free: 20702.0, maintenance: False, cpu-load: None, gateway: True}
MainThread::INFO::2016-06-14
08:15:44,452::brokerlink::111::ovirt_hosted_engine_ha.lib.brokerlink.BrokerLink::(notify)
Trying: notify time=1465906544.45 type=state_transition
detail=StartState-ReinitializeFSM hostname=njsevcnp02

On Tue, Jun 14, 2016 at 7:59 AM, Martin Sivak <msivak at redhat.com> wrote:

> Hi,
>
> is there anything interesting in the hosted engine log files?
> /var/log/ovirt-hosted-engine-ha/agent.log
>
> There should be something appearing there every 10 seconds or faster.
>
> What is the status of the hosted engine services? systemctl status
> ovirt-ha-agent ovirt-ha-broker
>
>
> Best regards
>
> --
> Martin Sivak
> SLA / oVirt
>
> On Sat, Jun 11, 2016 at 8:53 PM, Charles Kozler <charles at fixflyer.com>
> wrote:
> > See linked images please. As you can see all three nodes are reporting
> stale
> > data. The results of this are:
> >
> > 1. Not all VM's migrate seamlessly in the cluster. Sometimes I have to
> shut
> > them down to get them to be able to migrate again
> >
> > 2. Hosted engine refuses to move due to constraints (image). This part
> > doesnt make sense to me  because I can forcefully shut it down and then
> go
> > directly on a hosted engine node and bring it back up. Also, the Web UI
> > shows all nodes under the cluster except then it thinks its not apart of
> the
> > cluster
> >
> > 3. Time is in sync (image)
> >
> > 4. Storage is 100% fine. Gluster back end reports mirroring and status
> > 'started'. No split brain has occurred and ovirt nodes have never lost
> > connectivity to storage
> >
> > 5. I reinstalled all three nodes. For some reason only node 3 still
> shows as
> > having updates available. (image). For clarity, I did not click
> "upgrade" I
> > simply did 'reinstall' from the Web UI. Having looked at the output and
> > yum.log from /var/log it almost looks like it did do an update. All
> package
> > versions across all three nodes are the same (respective to ovirt/vdsm)
> > (image). For some reason though
> > ovirt-engine-appliance-3.6-20160126.1.el7.centos.noarch exists on node 1
> but
> > not on node 2 or 3. Could this be relative? I dont recall installing that
> > specifically on node 1 but I may have
> >
> > Been slamming my head on this so I am hoping you can provide some
> assistance
> >
> > http://imgur.com/a/6xkaS
> >
> > Thanks!
> >
> > --
> >
> > Charles Kozler
> > Vice President, IT Operations
> >
> > FIX Flyer, LLC
> > 225 Broadway | Suite 1600 | New York, NY 10007
> > 1-888-349-3593
> > http://www.fixflyer.com
> >
> > NOTICE TO RECIPIENT: THIS E-MAIL IS MEANT ONLY FOR THE INTENDED
> RECIPIENT(S)
> > OF THE TRANSMISSION, AND CONTAINS CONFIDENTIAL INFORMATION WHICH IS
> > PROPRIETARY TO FIX FLYER LLC.  ANY UNAUTHORIZED USE, COPYING,
> DISTRIBUTION,
> > OR DISSEMINATION IS STRICTLY PROHIBITED.  ALL RIGHTS TO THIS INFORMATION
> IS
> > RESERVED BY FIX FLYER LLC.  IF YOU ARE NOT THE INTENDED RECIPIENT, PLEASE
> > CONTACT THE SENDER BY REPLY E-MAIL AND PLEASE DELETE THIS E-MAIL FROM
> YOUR
> > SYSTEM AND DESTROY ANY COPIES.
> >
> > _______________________________________________
> > Users mailing list
> > Users at ovirt.org
> > http://lists.ovirt.org/mailman/listinfo/users
> >
>

-- 

*Charles Kozler*
*Vice President, IT Operations*

FIX Flyer, LLC
225 Broadway | Suite 1600 | New York, NY 10007
1-888-349-3593
http://www.fixflyer.com <http://fixflyer.com>

NOTICE TO RECIPIENT: THIS E-MAIL IS MEANT ONLY FOR THE INTENDED
RECIPIENT(S) OF THE TRANSMISSION, AND CONTAINS CONFIDENTIAL INFORMATION
WHICH IS PROPRIETARY TO FIX FLYER LLC.  ANY UNAUTHORIZED USE, COPYING,
DISTRIBUTION, OR DISSEMINATION IS STRICTLY PROHIBITED.  ALL RIGHTS TO THIS
INFORMATION IS RESERVED BY FIX FLYER LLC.  IF YOU ARE NOT THE INTENDED
RECIPIENT, PLEASE CONTACT THE SENDER BY REPLY E-MAIL AND PLEASE DELETE THIS
E-MAIL FROM YOUR SYSTEM AND DESTROY ANY COPIES.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20160614/c3d6877f/attachment-0001.html>