On 03/03/2014 05:27 PM, Martin Sivak wrote:
The agent is required. In fact it contains all the logic.
Thanks a lot for the information. I patched the agent and now it is running fine: # service ovirt-ha-agent status ovirt-ha-agent (pid 12385) is running... Btw, after starting the ha-agent the broken lock-file was fixed, too. So hosted engine is working fine now. Martin, Didi, thanks a lot for your help! Regards, René
-- Martin Sivák msivak@redhat.com Red Hat Czech RHEV-M SLA / Brno, CZ
----- Original Message -----
On 03/03/2014 02:05 PM, Martin Sivak wrote:
Hi René,
# python --version Python 2.6.6
Then I guess the traceback is my fault...
See http://gerrit.ovirt.org/#/c/25269/ for the fix. I will try to get it into the soonest release possible.
Thanks. Do I have to patch the files manually or is ovirt-ha-agent not strictly required for hosted engine? Some features like restarting engine on 2nd node want work if ovirt-ha-agent isn't working, I guess.
I can't see a full filesystem here:
Me neither. Is everything Read-Write? Read-Only FS might report no space left as well in some cases. Other than that, I do not know.
No, I can write to all disks. Btw, the same error message occurs on both nodes...
Regards, René
Regards -- Martin Sivák msivak@redhat.com Red Hat Czech RHEV-M SLA / Brno, CZ
----- Original Message -----
On 03/03/2014 12:05 PM, Martin Sivak wrote:
Hi René,
thanks for the report.
> TypeError: super() argument 1 must be type, not classobj What Python version are you using?
# python --version Python 2.6.6
You can debug a crash of this version of ha-agent using:
/usr/share/ovirt-hosted-engine-ha/ovirt-ha-agent --no-daemon --pdb
This gives me the same information as in vdsm.log
But this exception is trying to tell you that FSMLoggerAdapter(logging.LoggerAdapter) does not have object in the ancestor list. And that is very weird.
It can be related to the disk space issues.
> libvirtError: Failed to acquire lock: No space left on device
Check the free space on all your devices, including /tmp and /var. Or post the output of "df -h" command here
I can't see a full filesystem here:
# df -h Filesystem Size Used Avail Use% Mounted on /dev/mapper/vg0-lv_root 5.0G 1.1G 3.6G 24% / tmpfs 16G 0 16G 0% /dev/shm /dev/sda1 243M 45M 185M 20% /boot /dev/mapper/vg0-lv_data 281G 21G 261G 8% /data /dev/mapper/vg0-lv_tmp 2.0G 69M 1.9G 4% /tmp /dev/mapper/vg0-lv_var 5.0G 384M 4.3G 9% /var ovirt-host01:/engine 281G 21G 261G 8% /rhev/data-center/mnt/ovirt-host01:_engine
Thanks, René
Regards
-- Martin Sivák msivak@redhat.com Red Hat Czech RHEV-M SLA / Brno, CZ
----- Original Message -----
Il 03/03/2014 11:33, René Koch ha scritto: > Hi, > > I have some issues with hosted engine (oVirt 3.4 prerelease repo on > CentOS > 6.5). > My setups is the following: > 2 hosts (will be 4 in the future) with 4 GlusterFS shares: > - engine (for hosted engine) > - iso (for ISO domain) > - ovirt (oVirt storage domain) > > I had a split-brain situation today (after rebooting both nodes) on > hosted-engine.lockspace file on engine GlusterFS volume which I > resolved.
How did you solved it? By switching to NFS only?
> hosted engine used engine share via NFS (TCP) as glusterfs isn't > supported > for oVirt hosted engine, yet. I'll switch to GlusterFS as soon as oVirt > will support it (I hope this will be soon as RHEV 3.3 is already > supporting > GlusterFS for hosted engine). > > > First of all ovirt-ha-agent fails to start on both nodes: > > # service ovirt-ha-agent start > Starting ovirt-ha-agent: [ OK ] > # service ovirt-ha-agent status > ovirt-ha-agent dead but subsys locked > > > MainThread::INFO::2014-03-03 > 11:20:39,539::agent::52::ovirt_hosted_engine_ha.agent.agent.Agent::(run) > ovirt-hosted-engine-ha agent 1.1.0 started > MainThread::INFO::2014-03-03 > 11:20:39,590::hosted_engine::223::ovirt_hosted_engine_ha.agent.hosted_engine.HostedEngine::(_get_hostname) > Found > certificate common name: 10.0.200.101 > MainThread::CRITICAL::2014-03-03 > 11:20:39,590::agent::103::ovirt_hosted_engine_ha.agent.agent.Agent::(run) > Could not start ha-agent > Traceback (most recent call last): > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/agent/agent.py", > line 97, in run > self._run_agent() > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/agent/agent.py", > line 154, in _run_agent > hosted_engine.HostedEngine(self.shutdown_requested).start_monitoring() > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/agent/hosted_engine.py", > line 152, in __init__ > "STOP_VM": self._stop_engine_vm > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/agent/state_machine.py", > line 56, in __init__ > logger, actions) > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/lib/fsm/machine.py", > line 93, in __init__ > self._logger = FSMLoggerAdapter(logger, self) > File > "/usr/lib/python2.6/site-packages/ovirt_hosted_engine_ha/lib/fsm/machine.py", > line 16, in __init__ > super(FSMLoggerAdapter, self).__init__(logger, None) > TypeError: super() argument 1 must be type, not classobj > > > > If I want to start my hosted engine, I receive the following error in > vdsm > logs, which makes absolutly no sense to me, as there is plenty of disk > space available: > > Thread-62::DEBUG::2014-03-03 > 11:24:46,282::libvirtconnection::124::root::(wrapper) Unknown > libvirterror: ecode: 38 edom: 42 level: 2 message: Failed > to acquire lock: No space left on device
seems like a vdsm failure in starting monitor the hosted engine storage domain. Can you attach vdsm logs?
> Thread-62::DEBUG::2014-03-03 > 11:24:46,282::vm::2252::vm.Vm::(_startUnderlyingVm) > vmId=`f26dd37e-13b5-430c-b2f2-ecd098b82a91`::_ongoingCreations released > Thread-62::ERROR::2014-03-03 > 11:24:46,283::vm::2278::vm.Vm::(_startUnderlyingVm) > vmId=`f26dd37e-13b5-430c-b2f2-ecd098b82a91`::The vm start process > failed > Traceback (most recent call last): > File "/usr/share/vdsm/vm.py", line 2238, in _startUnderlyingVm > self._run() > File "/usr/share/vdsm/vm.py", line 3159, in _run > self._connection.createXML(domxml, flags), > File > "/usr/lib64/python2.6/site-packages/vdsm/libvirtconnection.py", > line > 92, in wrapper > ret = f(*args, **kwargs) > File "/usr/lib64/python2.6/site-packages/libvirt.py", line 2665, in > createXML > if ret is None:raise libvirtError('virDomainCreateXML() failed', > conn=self) > libvirtError: Failed to acquire lock: No space left on device > Thread-62::DEBUG::2014-03-03 > 11:24:46,286::vm::2720::vm.Vm::(setDownStatus) > vmId=`f26dd37e-13b5-430c-b2f2-ecd098b82a91`::Changed state to Down: > Failed > to acquire lock: No space left on device > > # df -h | grep engine > ovirt-host01:/engine 281G 21G 261G 8% > /rhev/data-center/mnt/ovirt-host01:_engine > > # sudo -u vdsm dd if=/dev/zero > of=/rhev/data-center/mnt/ovirt-host01:_engine/2851af27-8744-445d-9fb1-a0d083c8dc82/images/0e4d270f-2f7e-4b2b-847f-f114a4ba9bdc/test > bs=512 count=100 > 100+0 records in > 100+0 records out > 51200 bytes (51 kB) copied, 0.0230566 s, 2.2 MB/s > > > Could you give me some information on how to fix the ovirt-ha-agent and > then hosted-engine storage issue? Thanks a lot. > > Btw, I had some issues during installation which I will explain in > separate > emails. > >
-- Sandro Bonazzola Better technology. Faster innovation. Powered by community collaboration. See how it works at redhat.com