Thanks, will give that a go Karma++ :)
On Thu, Dec 15, 2016 at 12:13 PM, Simone Tiraboschi <stira...@redhat.com> wrote: > > > On Thu, Dec 15, 2016 at 12:04 PM, Renout Gerrits <m...@renout.nl> wrote: > >> Hi Simone, >> >> Do you mean the following? >> >> - create a new cluster with version 3.6. >> - Migrate HE to new cluster >> - Shutdown all VM's in old cluster >> - change compatibility version of old cluster to 3.6 >> - migrate HE back to old old cluster >> >> In this case the old cluster still thinks that the HE is running in it >> due to the ChangeVmCluster action that fails. From what I see this is fixed >> in 4.02: https://bugzilla.redhat.com/show_bug.cgi?id=1351533 >> But I can't to upgrade 4 yet. Do you know if this fix has been back >> ported to 3.6? >> > > Yes, it has been backported to 3.6.9: > https://gerrit.ovirt.org/#/c/63377/2 > > Another option is just to add a new hosted-engine host to the new 3.6 > cluster and restart the engine VM there from the hosted-engine CLI. > > For me it's hard to just try as I will need a maintenance window to >> shutdown all vm's. >> >> Or do you mean something completely different? >> >> Thanks, >> Renout >> >> On Thu, Dec 15, 2016 at 11:01 AM, Simone Tiraboschi <stira...@redhat.com> >> wrote: >> >>> >>> >>> On Thu, Dec 15, 2016 at 10:33 AM, Renout Gerrits <m...@renout.nl> wrote: >>> >>>> Hi All, >>>> >>>> We have an environment which we want to upgrade to ovirt 4.0. This was >>>> initially installed at 3.5, then upgraded to 3.6. >>>> Problem we're facing is that for an upgrade to 4.0 a compatibility >>>> version of 3.6 is required. >>>> When changing the cluster compatibility version of the 'Default' >>>> cluster from 3.5 to 3.6 we get the error in the gui: "Cannot change cluster >>>> compatibility version when a VM is active. please shutdown all VMs in the >>>> cluster." >>>> Even when we shutdown all vm's, except for the Hosted Engine we get >>>> this error. >>>> On the hosts a 'vdsClient -s 0 list' is done which will return the HE. >>>> In the engine logs we have the following error: "2016-12-08 >>>> 13:00:18,139 WARN [org.ovirt.engine.core.bll.st >>>> orage.UpdateStoragePoolCommand] (default task-25) [77a50037] >>>> CanDoAction of action 'UpdateStoragePool' failed for user admin@internal. >>>> Reasons: VAR__TYPE__STORAGE__POOL,VAR__ACTION__UPDATE,$ClustersList >>>> Default,ERROR_CANNOT_UPDATE_STORAGE_POOL_COMPATIBILITY_VERSI >>>> ON_BIGGER_THAN_CLUSTERS" >>>> >>>> So problem would be that the HE is in the Default cluster. But how does >>>> one change the compatibility version when the HE is down? >>>> I've tried shutting down the engine, changing the version in the DB: >>>> "UPDATE vds_groups SET compatibility_version='3.6';" and starting the >>>> engine again. >>>> >>>> When I do that and try to start a VM: >>>> 2016-12-09T13:30:21.346740Z qemu-kvm: warning: CPU(s) not present in >>>> any NUMA nodes: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >>>> 2016-12-09T13:30:21.346883Z qemu-kvm: warning: All CPU(s) up to maxcpus >>>> should be described in NUMA config >>>> 2016-12-09T13:30:21.355699Z qemu-kvm: "-memory 'slots|maxmem'" is not >>>> supported by: rhel6.5.0 >>>> >>>> So that change was rolled back to compatibilty 3.5. After that we we're >>>> able to start vm's again. >>>> Please note that all hosts and HE are EL7. >>>> >>>> To me this doesn't seem like a strange set-up or upgrade path. Would it >>>> be possible to start the HE in another cluster than Default or is there a >>>> way to bypass the vdsClient list check? >>>> What is the recommended way of upgrading the HE in this case? >>>> >>> >>> Please take a look here: >>> https://bugzilla.redhat.com/show_bug.cgi?id=1364557 >>> >>> >>>> >>>> Kind regards, >>>> Renout >>>> >>>> _______________________________________________ >>>> 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