Re: [ovirt-devel] Re: URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
we have a passing ovirt-engine build today. Thank you all for a fast response. Dafna On Thu, May 9, 2019 at 12:43 PM Sandro Bonazzola wrote: > > > Il giorno gio 9 mag 2019 alle ore 12:59 Dafna Ron ha > scritto: > >> As IL are on independence day, anyone else can merge? >> https://gerrit.ovirt.org/#/c/99845/ >> >> > I have merge rights but I need at least CI to pass. Waiting on jenkins. > > >> >> On Thu, May 9, 2019 at 11:30 AM Dafna Ron wrote: >> >>> Thanks Andrej. >>> I will follow the patch and update. >>> Dafna >>> >>> On Thu, May 9, 2019 at 11:23 AM Andrej Krejcir >>> wrote: >>> Hi, Ok, I have posted the reverting patch: https://gerrit.ovirt.org/#/c/99845/ I'm still investigating what is the problem. Sorry for the delay, we had a public holiday yesturday. Andrej On Thu, 9 May 2019 at 11:20, Dafna Ron wrote: > Hi, > > I have not heard back on this issue and ovirt-engine has been broken > for the past 3 days. > > As this does not seem a simple debug and fix I suggest reverting the > patch and investigating later. > > thanks, > Dafna > > > > On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: > >> Any news? >> >> Thanks, >> Dafna >> >> >> On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: >> >>> thanks for the quick reply and investigation. >>> Please update me if I can help any further and if you find the cause >>> and have a patch let me know. >>> Note that ovirt-engine project is broken and if we cannot find the >>> cause relatively fast we should consider reverting the patch to allow a >>> new >>> package to be built in CQ with other changes that were submitted. >>> >>> Thanks, >>> Dafna >>> >>> >>> On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir >>> wrote: >>> After running a few OSTs manually, it seems that the patch is the cause. Investigating... On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: > Hi, > > The issue is probably not caused by the patch. > > This log line means that the VM does not exist in the DB: > > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] > Validation > of action 'MigrateMultipleVms' failed for user admin@internal-authz. > Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND > > I will investigate more, why the VM is missing. > > On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > >> Hi, >> >> We are failing test upgrade_hosts on >> upgrade-from-release-suite-master. >> From the logs I can see that we are calling migrate vm when we >> have only one host and the vm seem to have been shut down before the >> maintenance call is issued. >> >> Can you please look into this? >> >> suspected patch reported as root cause by CQ is: >> >> https://gerrit.ovirt.org/#/c/98920/ - core: Add >> MigrateMultipleVms command and use it for host maintenance >> >> >> logs are found here: >> >> >> >> http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ >> >> >> I can see the issue is vm migration when putting host in >> maintenance: >> >> >> 2019-05-07 06:02:04,170-04 INFO >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) >> [05592db2-f859-487b-b779-4b32eec5bab >> 3] Running command: MaintenanceVdsCommand internal: true. >> Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: >> VDS >> 2019-05-07 06:02:04,215-04 WARN >> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >> Validation >> of action >> 'MigrateMultipleVms' failed for user admin@internal-authz. >> Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND >> 2019-05-07 06:02:04,221-04 ERROR >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >> Failed to >> migrate one or >> more VMs. >> 2019-05-07 06:02:04,227-04 ERROR >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN >> T_ID: VDS_MAINTENANCE_FAILED(17), Failed to s
Re: [ovirt-devel] Re: URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Il giorno gio 9 mag 2019 alle ore 12:59 Dafna Ron ha scritto: > As IL are on independence day, anyone else can merge? > https://gerrit.ovirt.org/#/c/99845/ > > I have merge rights but I need at least CI to pass. Waiting on jenkins. > > On Thu, May 9, 2019 at 11:30 AM Dafna Ron wrote: > >> Thanks Andrej. >> I will follow the patch and update. >> Dafna >> >> On Thu, May 9, 2019 at 11:23 AM Andrej Krejcir >> wrote: >> >>> Hi, >>> >>> Ok, I have posted the reverting patch: >>> https://gerrit.ovirt.org/#/c/99845/ >>> >>> I'm still investigating what is the problem. Sorry for the delay, we had >>> a public holiday yesturday. >>> >>> >>> Andrej >>> >>> On Thu, 9 May 2019 at 11:20, Dafna Ron wrote: >>> Hi, I have not heard back on this issue and ovirt-engine has been broken for the past 3 days. As this does not seem a simple debug and fix I suggest reverting the patch and investigating later. thanks, Dafna On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: > Any news? > > Thanks, > Dafna > > > On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: > >> thanks for the quick reply and investigation. >> Please update me if I can help any further and if you find the cause >> and have a patch let me know. >> Note that ovirt-engine project is broken and if we cannot find the >> cause relatively fast we should consider reverting the patch to allow a >> new >> package to be built in CQ with other changes that were submitted. >> >> Thanks, >> Dafna >> >> >> On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir >> wrote: >> >>> After running a few OSTs manually, it seems that the patch is the >>> cause. Investigating... >>> >>> On Tue, 7 May 2019 at 14:58, Andrej Krejcir >>> wrote: >>> Hi, The issue is probably not caused by the patch. This log line means that the VM does not exist in the DB: 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND I will investigate more, why the VM is missing. On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > Hi, > > We are failing test upgrade_hosts on > upgrade-from-release-suite-master. > From the logs I can see that we are calling migrate vm when we > have only one host and the vm seem to have been shut down before the > maintenance call is issued. > > Can you please look into this? > > suspected patch reported as root cause by CQ is: > > https://gerrit.ovirt.org/#/c/98920/ - core: Add > MigrateMultipleVms command and use it for host maintenance > > > logs are found here: > > > > http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ > > > I can see the issue is vm migration when putting host in > maintenance: > > > 2019-05-07 06:02:04,170-04 INFO > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) > [05592db2-f859-487b-b779-4b32eec5bab > 3] Running command: MaintenanceVdsCommand internal: true. Entities > affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] > Validation > of action > 'MigrateMultipleVms' failed for user admin@internal-authz. > Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND > 2019-05-07 06:02:04,221-04 ERROR > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] > Failed to > migrate one or > more VMs. > 2019-05-07 06:02:04,227-04 ERROR > [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN > T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host > lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. > 2019-05-07 06:02:04,239-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock > Acquired to object 'Eng >>>
Re: URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
As IL are on independence day, anyone else can merge? https://gerrit.ovirt.org/#/c/99845/ On Thu, May 9, 2019 at 11:30 AM Dafna Ron wrote: > Thanks Andrej. > I will follow the patch and update. > Dafna > > On Thu, May 9, 2019 at 11:23 AM Andrej Krejcir > wrote: > >> Hi, >> >> Ok, I have posted the reverting patch: >> https://gerrit.ovirt.org/#/c/99845/ >> >> I'm still investigating what is the problem. Sorry for the delay, we had >> a public holiday yesturday. >> >> >> Andrej >> >> On Thu, 9 May 2019 at 11:20, Dafna Ron wrote: >> >>> Hi, >>> >>> I have not heard back on this issue and ovirt-engine has been broken for >>> the past 3 days. >>> >>> As this does not seem a simple debug and fix I suggest reverting the >>> patch and investigating later. >>> >>> thanks, >>> Dafna >>> >>> >>> >>> On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: >>> Any news? Thanks, Dafna On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: > thanks for the quick reply and investigation. > Please update me if I can help any further and if you find the cause > and have a patch let me know. > Note that ovirt-engine project is broken and if we cannot find the > cause relatively fast we should consider reverting the patch to allow a > new > package to be built in CQ with other changes that were submitted. > > Thanks, > Dafna > > > On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir > wrote: > >> After running a few OSTs manually, it seems that the patch is the >> cause. Investigating... >> >> On Tue, 7 May 2019 at 14:58, Andrej Krejcir >> wrote: >> >>> Hi, >>> >>> The issue is probably not caused by the patch. >>> >>> This log line means that the VM does not exist in the DB: >>> >>> 2019-05-07 06:02:04,215-04 WARN >>> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >>> Validation >>> of action 'MigrateMultipleVms' failed for user admin@internal-authz. >>> Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND >>> >>> I will investigate more, why the VM is missing. >>> >>> On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: >>> Hi, We are failing test upgrade_hosts on upgrade-from-release-suite-master. From the logs I can see that we are calling migrate vm when we have only one host and the vm seem to have been shut down before the maintenance call is issued. Can you please look into this? suspected patch reported as root cause by CQ is: https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms command and use it for host maintenance logs are found here: http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ I can see the issue is vm migration when putting host in maintenance: 2019-05-07 06:02:04,170-04 INFO [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [05592db2-f859-487b-b779-4b32eec5bab 3] Running command: MaintenanceVdsCommand internal: true. Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND 2019-05-07 06:02:04,221-04 ERROR [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to migrate one or more VMs. 2019-05-07 06:02:04,227-04 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. 2019-05-07 06:02:04,239-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock Acquired to object 'Eng ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', sharedLocks=''}' 2019-05-07 06:02:04,242-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] >
Re: URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Thanks Andrej. I will follow the patch and update. Dafna On Thu, May 9, 2019 at 11:23 AM Andrej Krejcir wrote: > Hi, > > Ok, I have posted the reverting patch: https://gerrit.ovirt.org/#/c/99845/ > > I'm still investigating what is the problem. Sorry for the delay, we had a > public holiday yesturday. > > > Andrej > > On Thu, 9 May 2019 at 11:20, Dafna Ron wrote: > >> Hi, >> >> I have not heard back on this issue and ovirt-engine has been broken for >> the past 3 days. >> >> As this does not seem a simple debug and fix I suggest reverting the >> patch and investigating later. >> >> thanks, >> Dafna >> >> >> >> On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: >> >>> Any news? >>> >>> Thanks, >>> Dafna >>> >>> >>> On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: >>> thanks for the quick reply and investigation. Please update me if I can help any further and if you find the cause and have a patch let me know. Note that ovirt-engine project is broken and if we cannot find the cause relatively fast we should consider reverting the patch to allow a new package to be built in CQ with other changes that were submitted. Thanks, Dafna On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir wrote: > After running a few OSTs manually, it seems that the patch is the > cause. Investigating... > > On Tue, 7 May 2019 at 14:58, Andrej Krejcir > wrote: > >> Hi, >> >> The issue is probably not caused by the patch. >> >> This log line means that the VM does not exist in the DB: >> >> 2019-05-07 06:02:04,215-04 WARN >> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >> Validation >> of action 'MigrateMultipleVms' failed for user admin@internal-authz. >> Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND >> >> I will investigate more, why the VM is missing. >> >> On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: >> >>> Hi, >>> >>> We are failing test upgrade_hosts on >>> upgrade-from-release-suite-master. >>> From the logs I can see that we are calling migrate vm when we have >>> only one host and the vm seem to have been shut down before the >>> maintenance >>> call is issued. >>> >>> Can you please look into this? >>> >>> suspected patch reported as root cause by CQ is: >>> >>> https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms >>> command and use it for host maintenance >>> >>> >>> logs are found here: >>> >>> >>> >>> http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ >>> >>> >>> I can see the issue is vm migration when putting host in >>> maintenance: >>> >>> >>> 2019-05-07 06:02:04,170-04 INFO >>> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) >>> [05592db2-f859-487b-b779-4b32eec5bab >>> 3] Running command: MaintenanceVdsCommand internal: true. Entities >>> affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS >>> 2019-05-07 06:02:04,215-04 WARN >>> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >>> Validation >>> of action >>> 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: >>> ACTION_TYPE_FAILED_VMS_NOT_FOUND >>> 2019-05-07 06:02:04,221-04 ERROR >>> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed >>> to >>> migrate one or >>> more VMs. >>> 2019-05-07 06:02:04,227-04 ERROR >>> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN >>> T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host >>> lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. >>> 2019-05-07 06:02:04,239-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >>> Acquired to object 'Eng >>> ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', >>> sharedLocks=''}' >>> 2019-05-07 06:02:04,242-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running >>> command: ActivateVds >>> Command internal: true. Entities affected : ID: >>> 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group >>> MANIPULATE_HOST >>> with role type ADMIN >>> 2019-05-07 06:02:04,243-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsComm
Re: URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Hi, Ok, I have posted the reverting patch: https://gerrit.ovirt.org/#/c/99845/ I'm still investigating what is the problem. Sorry for the delay, we had a public holiday yesturday. Andrej On Thu, 9 May 2019 at 11:20, Dafna Ron wrote: > Hi, > > I have not heard back on this issue and ovirt-engine has been broken for > the past 3 days. > > As this does not seem a simple debug and fix I suggest reverting the patch > and investigating later. > > thanks, > Dafna > > > > On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: > >> Any news? >> >> Thanks, >> Dafna >> >> >> On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: >> >>> thanks for the quick reply and investigation. >>> Please update me if I can help any further and if you find the cause and >>> have a patch let me know. >>> Note that ovirt-engine project is broken and if we cannot find the cause >>> relatively fast we should consider reverting the patch to allow a new >>> package to be built in CQ with other changes that were submitted. >>> >>> Thanks, >>> Dafna >>> >>> >>> On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir >>> wrote: >>> After running a few OSTs manually, it seems that the patch is the cause. Investigating... On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: > Hi, > > The issue is probably not caused by the patch. > > This log line means that the VM does not exist in the DB: > > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] > Validation > of action 'MigrateMultipleVms' failed for user admin@internal-authz. > Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND > > I will investigate more, why the VM is missing. > > On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > >> Hi, >> >> We are failing test upgrade_hosts on >> upgrade-from-release-suite-master. >> From the logs I can see that we are calling migrate vm when we have >> only one host and the vm seem to have been shut down before the >> maintenance >> call is issued. >> >> Can you please look into this? >> >> suspected patch reported as root cause by CQ is: >> >> https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms >> command and use it for host maintenance >> >> >> logs are found here: >> >> >> >> http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ >> >> >> I can see the issue is vm migration when putting host in maintenance: >> >> >> 2019-05-07 06:02:04,170-04 INFO >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) >> [05592db2-f859-487b-b779-4b32eec5bab >> 3] Running command: MaintenanceVdsCommand internal: true. Entities >> affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS >> 2019-05-07 06:02:04,215-04 WARN >> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] >> Validation >> of action >> 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: >> ACTION_TYPE_FAILED_VMS_NOT_FOUND >> 2019-05-07 06:02:04,221-04 ERROR >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed >> to >> migrate one or >> more VMs. >> 2019-05-07 06:02:04,227-04 ERROR >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN >> T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host >> lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. >> 2019-05-07 06:02:04,239-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >> Acquired to object 'Eng >> ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', >> sharedLocks=''}' >> 2019-05-07 06:02:04,242-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running >> command: ActivateVds >> Command internal: true. Entities affected : ID: >> 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group >> MANIPULATE_HOST >> with role type ADMIN >> 2019-05-07 06:02:04,243-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before >> acquiring lock in ord >> er to prevent monitoring for host >> 'lago-upgrade-from-release-suite-master-host-0' from data-center >> 'test-dc' >> 2019-05-07 06:02
URGENT - ovirt-engine broken for 3 days Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Hi, I have not heard back on this issue and ovirt-engine has been broken for the past 3 days. As this does not seem a simple debug and fix I suggest reverting the patch and investigating later. thanks, Dafna On Wed, May 8, 2019 at 9:42 AM Dafna Ron wrote: > Any news? > > Thanks, > Dafna > > > On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: > >> thanks for the quick reply and investigation. >> Please update me if I can help any further and if you find the cause and >> have a patch let me know. >> Note that ovirt-engine project is broken and if we cannot find the cause >> relatively fast we should consider reverting the patch to allow a new >> package to be built in CQ with other changes that were submitted. >> >> Thanks, >> Dafna >> >> >> On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir >> wrote: >> >>> After running a few OSTs manually, it seems that the patch is the cause. >>> Investigating... >>> >>> On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: >>> Hi, The issue is probably not caused by the patch. This log line means that the VM does not exist in the DB: 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND I will investigate more, why the VM is missing. On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > Hi, > > We are failing test upgrade_hosts on > upgrade-from-release-suite-master. > From the logs I can see that we are calling migrate vm when we have > only one host and the vm seem to have been shut down before the > maintenance > call is issued. > > Can you please look into this? > > suspected patch reported as root cause by CQ is: > > https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms > command and use it for host maintenance > > > logs are found here: > > > > http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ > > > I can see the issue is vm migration when putting host in maintenance: > > > 2019-05-07 06:02:04,170-04 INFO > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) > [05592db2-f859-487b-b779-4b32eec5bab > 3] Running command: MaintenanceVdsCommand internal: true. Entities > affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] > Validation > of action > 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: > ACTION_TYPE_FAILED_VMS_NOT_FOUND > 2019-05-07 06:02:04,221-04 ERROR > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to > migrate one or > more VMs. > 2019-05-07 06:02:04,227-04 ERROR > [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN > T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host > lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. > 2019-05-07 06:02:04,239-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock > Acquired to object 'Eng > ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', > sharedLocks=''}' > 2019-05-07 06:02:04,242-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running > command: ActivateVds > Command internal: true. Entities affected : ID: > 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST > with role type ADMIN > 2019-05-07 06:02:04,243-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before > acquiring lock in ord > er to prevent monitoring for host > 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' > 2019-05-07 06:02:04,243-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock > acquired, from now a mo > nitoring of host will be skipped for host > 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' > 2019-05-07 06:02:04,252-04 INFO > [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] > (
Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Any news? Thanks, Dafna On Tue, May 7, 2019 at 4:57 PM Dafna Ron wrote: > thanks for the quick reply and investigation. > Please update me if I can help any further and if you find the cause and > have a patch let me know. > Note that ovirt-engine project is broken and if we cannot find the cause > relatively fast we should consider reverting the patch to allow a new > package to be built in CQ with other changes that were submitted. > > Thanks, > Dafna > > > On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir wrote: > >> After running a few OSTs manually, it seems that the patch is the cause. >> Investigating... >> >> On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: >> >>> Hi, >>> >>> The issue is probably not caused by the patch. >>> >>> This log line means that the VM does not exist in the DB: >>> >>> 2019-05-07 06:02:04,215-04 WARN >>> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation >>> of action 'MigrateMultipleVms' failed for user admin@internal-authz. >>> Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND >>> >>> I will investigate more, why the VM is missing. >>> >>> On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: >>> Hi, We are failing test upgrade_hosts on upgrade-from-release-suite-master. From the logs I can see that we are calling migrate vm when we have only one host and the vm seem to have been shut down before the maintenance call is issued. Can you please look into this? suspected patch reported as root cause by CQ is: https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms command and use it for host maintenance logs are found here: http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ I can see the issue is vm migration when putting host in maintenance: 2019-05-07 06:02:04,170-04 INFO [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [05592db2-f859-487b-b779-4b32eec5bab 3] Running command: MaintenanceVdsCommand internal: true. Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND 2019-05-07 06:02:04,221-04 ERROR [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to migrate one or more VMs. 2019-05-07 06:02:04,227-04 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. 2019-05-07 06:02:04,239-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock Acquired to object 'Eng ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', sharedLocks=''}' 2019-05-07 06:02:04,242-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running command: ActivateVds Command internal: true. Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST with role type ADMIN 2019-05-07 06:02:04,243-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before acquiring lock in ord er to prevent monitoring for host 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' 2019-05-07 06:02:04,243-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock acquired, from now a mo nitoring of host will be skipped for host 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' 2019-05-07 06:02:04,252-04 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] START, SetVdsStatu sVDSCommand(HostName = lago-upgrade-from-release-suite-master-host-0, SetVdsStatusVDSCommandParameters:{hostId='38e1379b-c3b6-4a2e-91df-d1f346e414a9', status='Unassigned', n onOperationalReason='NONE', stopSpmFailureLogged='false', maintenanceReason='null'}), log id: 2c8aa211 2019-05-07 06
Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
thanks for the quick reply and investigation. Please update me if I can help any further and if you find the cause and have a patch let me know. Note that ovirt-engine project is broken and if we cannot find the cause relatively fast we should consider reverting the patch to allow a new package to be built in CQ with other changes that were submitted. Thanks, Dafna On Tue, May 7, 2019 at 4:42 PM Andrej Krejcir wrote: > After running a few OSTs manually, it seems that the patch is the cause. > Investigating... > > On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: > >> Hi, >> >> The issue is probably not caused by the patch. >> >> This log line means that the VM does not exist in the DB: >> >> 2019-05-07 06:02:04,215-04 WARN >> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation >> of action 'MigrateMultipleVms' failed for user admin@internal-authz. >> Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND >> >> I will investigate more, why the VM is missing. >> >> On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: >> >>> Hi, >>> >>> We are failing test upgrade_hosts on upgrade-from-release-suite-master. >>> From the logs I can see that we are calling migrate vm when we have only >>> one host and the vm seem to have been shut down before the maintenance call >>> is issued. >>> >>> Can you please look into this? >>> >>> suspected patch reported as root cause by CQ is: >>> >>> https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms >>> command and use it for host maintenance >>> >>> >>> logs are found here: >>> >>> >>> >>> http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ >>> >>> >>> I can see the issue is vm migration when putting host in maintenance: >>> >>> >>> 2019-05-07 06:02:04,170-04 INFO >>> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) >>> [05592db2-f859-487b-b779-4b32eec5bab >>> 3] Running command: MaintenanceVdsCommand internal: true. Entities >>> affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS >>> 2019-05-07 06:02:04,215-04 WARN >>> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation >>> of action >>> 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: >>> ACTION_TYPE_FAILED_VMS_NOT_FOUND >>> 2019-05-07 06:02:04,221-04 ERROR >>> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to >>> migrate one or >>> more VMs. >>> 2019-05-07 06:02:04,227-04 ERROR >>> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN >>> T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host >>> lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. >>> 2019-05-07 06:02:04,239-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >>> Acquired to object 'Eng >>> ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', >>> sharedLocks=''}' >>> 2019-05-07 06:02:04,242-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running >>> command: ActivateVds >>> Command internal: true. Entities affected : ID: >>> 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST >>> with role type ADMIN >>> 2019-05-07 06:02:04,243-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before >>> acquiring lock in ord >>> er to prevent monitoring for host >>> 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' >>> 2019-05-07 06:02:04,243-04 INFO >>> [org.ovirt.engine.core.bll.ActivateVdsCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >>> acquired, from now a mo >>> nitoring of host will be skipped for host >>> 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' >>> 2019-05-07 06:02:04,252-04 INFO >>> [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] START, >>> SetVdsStatu >>> sVDSCommand(HostName = lago-upgrade-from-release-suite-master-host-0, >>> SetVdsStatusVDSCommandParameters:{hostId='38e1379b-c3b6-4a2e-91df-d1f346e414a9', >>> status='Unassigned', n >>> onOperationalReason='NONE', stopSpmFailureLogged='false', >>> maintenanceReason='null'}), log id: 2c8aa211 >>> 2019-05-07 06:02:04,256-04 INFO >>> [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] >>> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] FINISH, >>> SetVdsStat >>> usVDSCommand, return: , lo
Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
After running a few OSTs manually, it seems that the patch is the cause. Investigating... On Tue, 7 May 2019 at 14:58, Andrej Krejcir wrote: > Hi, > > The issue is probably not caused by the patch. > > This log line means that the VM does not exist in the DB: > > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation > of action 'MigrateMultipleVms' failed for user admin@internal-authz. > Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND > > I will investigate more, why the VM is missing. > > On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > >> Hi, >> >> We are failing test upgrade_hosts on upgrade-from-release-suite-master. >> From the logs I can see that we are calling migrate vm when we have only >> one host and the vm seem to have been shut down before the maintenance call >> is issued. >> >> Can you please look into this? >> >> suspected patch reported as root cause by CQ is: >> >> https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms >> command and use it for host maintenance >> >> >> logs are found here: >> >> >> >> http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ >> >> >> I can see the issue is vm migration when putting host in maintenance: >> >> >> 2019-05-07 06:02:04,170-04 INFO >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) >> [05592db2-f859-487b-b779-4b32eec5bab >> 3] Running command: MaintenanceVdsCommand internal: true. Entities >> affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS >> 2019-05-07 06:02:04,215-04 WARN >> [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation >> of action >> 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: >> ACTION_TYPE_FAILED_VMS_NOT_FOUND >> 2019-05-07 06:02:04,221-04 ERROR >> [org.ovirt.engine.core.bll.MaintenanceVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to >> migrate one or >> more VMs. >> 2019-05-07 06:02:04,227-04 ERROR >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN >> T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host >> lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. >> 2019-05-07 06:02:04,239-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >> Acquired to object 'Eng >> ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', >> sharedLocks=''}' >> 2019-05-07 06:02:04,242-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running >> command: ActivateVds >> Command internal: true. Entities affected : ID: >> 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST >> with role type ADMIN >> 2019-05-07 06:02:04,243-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before >> acquiring lock in ord >> er to prevent monitoring for host >> 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' >> 2019-05-07 06:02:04,243-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock >> acquired, from now a mo >> nitoring of host will be skipped for host >> 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' >> 2019-05-07 06:02:04,252-04 INFO >> [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] START, >> SetVdsStatu >> sVDSCommand(HostName = lago-upgrade-from-release-suite-master-host-0, >> SetVdsStatusVDSCommandParameters:{hostId='38e1379b-c3b6-4a2e-91df-d1f346e414a9', >> status='Unassigned', n >> onOperationalReason='NONE', stopSpmFailureLogged='false', >> maintenanceReason='null'}), log id: 2c8aa211 >> 2019-05-07 06:02:04,256-04 INFO >> [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] FINISH, >> SetVdsStat >> usVDSCommand, return: , log id: 2c8aa211 >> 2019-05-07 06:02:04,261-04 INFO >> [org.ovirt.engine.core.bll.ActivateVdsCommand] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Activate >> host finished. Lock >> released. Monitoring can run now for host >> 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' >> 2019-05-07 06:02:04,265-04 INFO >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] >> (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] EVEN >> T_ID: VDS_ACTIVATE(16), Acti
Re: Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Hi, The issue is probably not caused by the patch. This log line means that the VM does not exist in the DB: 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND I will investigate more, why the VM is missing. On Tue, 7 May 2019 at 14:07, Dafna Ron wrote: > Hi, > > We are failing test upgrade_hosts on upgrade-from-release-suite-master. > From the logs I can see that we are calling migrate vm when we have only > one host and the vm seem to have been shut down before the maintenance call > is issued. > > Can you please look into this? > > suspected patch reported as root cause by CQ is: > > https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms > command and use it for host maintenance > > > logs are found here: > > > > http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ > > > I can see the issue is vm migration when putting host in maintenance: > > > 2019-05-07 06:02:04,170-04 INFO > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) > [05592db2-f859-487b-b779-4b32eec5bab > 3] Running command: MaintenanceVdsCommand internal: true. Entities > affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS > 2019-05-07 06:02:04,215-04 WARN > [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation > of action > 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: > ACTION_TYPE_FAILED_VMS_NOT_FOUND > 2019-05-07 06:02:04,221-04 ERROR > [org.ovirt.engine.core.bll.MaintenanceVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to > migrate one or > more VMs. > 2019-05-07 06:02:04,227-04 ERROR > [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN > T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host > lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. > 2019-05-07 06:02:04,239-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock > Acquired to object 'Eng > ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', > sharedLocks=''}' > 2019-05-07 06:02:04,242-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running > command: ActivateVds > Command internal: true. Entities affected : ID: > 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST > with role type ADMIN > 2019-05-07 06:02:04,243-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before > acquiring lock in ord > er to prevent monitoring for host > 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' > 2019-05-07 06:02:04,243-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock > acquired, from now a mo > nitoring of host will be skipped for host > 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' > 2019-05-07 06:02:04,252-04 INFO > [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] START, > SetVdsStatu > sVDSCommand(HostName = lago-upgrade-from-release-suite-master-host-0, > SetVdsStatusVDSCommandParameters:{hostId='38e1379b-c3b6-4a2e-91df-d1f346e414a9', > status='Unassigned', n > onOperationalReason='NONE', stopSpmFailureLogged='false', > maintenanceReason='null'}), log id: 2c8aa211 > 2019-05-07 06:02:04,256-04 INFO > [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] FINISH, > SetVdsStat > usVDSCommand, return: , log id: 2c8aa211 > 2019-05-07 06:02:04,261-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Activate > host finished. Lock > released. Monitoring can run now for host > 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' > 2019-05-07 06:02:04,265-04 INFO > [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] EVEN > T_ID: VDS_ACTIVATE(16), Activation of host > lago-upgrade-from-release-suite-master-host-0 initiated by > admin@internal-authz. > 2019-05-07 06:02:04,266-04 INFO > [org.ovirt.engine.core.bll.ActivateVdsCommand] > (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock freed > to
Subject: [ OST Failure Report ] [ oVirt Master (ovirt-engine) ] [ 05-05-2019 ] [ upgrade_hosts ]
Hi, We are failing test upgrade_hosts on upgrade-from-release-suite-master. >From the logs I can see that we are calling migrate vm when we have only one host and the vm seem to have been shut down before the maintenance call is issued. Can you please look into this? suspected patch reported as root cause by CQ is: https://gerrit.ovirt.org/#/c/98920/ - core: Add MigrateMultipleVms command and use it for host maintenance logs are found here: http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/14021/artifact/upgrade-from-release-suite.el7.x86_64/test_logs/upgrade-from-release-suite-master/post-004_basic_sanity.py/ I can see the issue is vm migration when putting host in maintenance: 2019-05-07 06:02:04,170-04 INFO [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [05592db2-f859-487b-b779-4b32eec5bab 3] Running command: MaintenanceVdsCommand internal: true. Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDS 2019-05-07 06:02:04,215-04 WARN [org.ovirt.engine.core.bll.MigrateMultipleVmsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Validation of action 'MigrateMultipleVms' failed for user admin@internal-authz. Reasons: ACTION_TYPE_FAILED_VMS_NOT_FOUND 2019-05-07 06:02:04,221-04 ERROR [org.ovirt.engine.core.bll.MaintenanceVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] Failed to migrate one or more VMs. 2019-05-07 06:02:04,227-04 ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [33485140] EVEN T_ID: VDS_MAINTENANCE_FAILED(17), Failed to switch Host lago-upgrade-from-release-suite-master-host-0 to Maintenance mode. 2019-05-07 06:02:04,239-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock Acquired to object 'Eng ineLock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', sharedLocks=''}' 2019-05-07 06:02:04,242-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Running command: ActivateVds Command internal: true. Entities affected : ID: 38e1379b-c3b6-4a2e-91df-d1f346e414a9 Type: VDSAction group MANIPULATE_HOST with role type ADMIN 2019-05-07 06:02:04,243-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Before acquiring lock in ord er to prevent monitoring for host 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' 2019-05-07 06:02:04,243-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock acquired, from now a mo nitoring of host will be skipped for host 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' 2019-05-07 06:02:04,252-04 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] START, SetVdsStatu sVDSCommand(HostName = lago-upgrade-from-release-suite-master-host-0, SetVdsStatusVDSCommandParameters:{hostId='38e1379b-c3b6-4a2e-91df-d1f346e414a9', status='Unassigned', n onOperationalReason='NONE', stopSpmFailureLogged='false', maintenanceReason='null'}), log id: 2c8aa211 2019-05-07 06:02:04,256-04 INFO [org.ovirt.engine.core.vdsbroker.SetVdsStatusVDSCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] FINISH, SetVdsStat usVDSCommand, return: , log id: 2c8aa211 2019-05-07 06:02:04,261-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Activate host finished. Lock released. Monitoring can run now for host 'lago-upgrade-from-release-suite-master-host-0' from data-center 'test-dc' 2019-05-07 06:02:04,265-04 INFO [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] EVEN T_ID: VDS_ACTIVATE(16), Activation of host lago-upgrade-from-release-suite-master-host-0 initiated by admin@internal-authz. 2019-05-07 06:02:04,266-04 INFO [org.ovirt.engine.core.bll.ActivateVdsCommand] (EE-ManagedThreadFactory-commandCoordinator-Thread-2) [70840477] Lock freed to object 'Engine Lock:{exclusiveLocks='[38e1379b-c3b6-4a2e-91df-d1f346e414a9=VDS]', sharedLocks=''}' 2019-05-07 06:02:04,484-04 ERROR [org.ovirt.engine.core.bll.hostdeploy.HostUpgradeCallback] (EE-ManagedThreadFactory-engineScheduled-Thread-96) [05592db2-f859-487b-b779-4b32 eec5bab3] Host 'lago-upgrade-from-release-suite-master-host-0' failed to move to maintenance mode. Upgrade process is terminated. I can see there was only one vm running: drwxrwxr-x. 2 dron dron 1024 May 7 11:49 qemu [dron@dron post-004_basic_sanity.py]$ ls -l lago-upgrade-from-release-suite-master-host-0/_var_log/libvirt/qemu/ total 6 -rw-rw-r--. 1 dron dron 4466 May 7 10:12 vm-with-iface.log