[ovirt-users] Re: Change hostname of physical hosts under an oVirt and Gluster combination

Unassigned in 99% of the cases means engine cannot communicate with vdsm.service on the host. Check that vdsm.service & glusterd.service are running. Check gluster volume status for each volume. If they are OK - from UI just select activate.

Best Regards, Strahil NikolovOn Sep 25, 2019 22:24, TomK <tomkcpr(a)mdevsys.com&gt; wrote: > > On 9/25/2019 8:49 AM, TomK wrote: >> On 9/24/2019 10:55 AM, Sahina Bose wrote: >>> >>> >>> On Tue, Sep 24, 2019 at 6:38 PM TomK <tomkcpr(a)mdevsys.com >>> <mailto:tomkcpr@mdevsys.com>> wrote: >>> >>>     Hey Sahina, >>> >>>     Thanks very much. >>> >>>     I've taken a quick glance and it does mention I need a third host >>> which >>>     I don't have. >>> >>>     However, since I wrote, I removed everything.  Had to force remove >>>     hosts >>>     however the gluster setup and data in it stayed (Or so I'm thinking). >>> >>>     I'll read the article more thoroughly however is there a specific >>>     procedure for adding existing oVirt gluster volumes back into a >>> cluster? >>> >>>     One additional question.  Is the sequence below correct? >>> >>>     1) Create Data Center >>>     2) Create Cluster within the DC >>> >>> >>> When you create the cluster - there's an Import cluster option, if I >>> remember correctly. This should discover the peers and volumes and add >>> it to the engine >> >> Trying that now.  On a side note, the autodiscovery box is hard to edit. >> >> So in the meantime, trying to up the version to take advantage of some >> new features and ensure Gluster 6 is fully supported. Not having luck >> with the oVirt UI upgrade at the moment ( Not a self hosted engine. It's >> on a separate VM.  ): >> >> --> Processing Conflict: >> ovirt-engine-setup-plugin-ovirt-engine-4.3.5.5-1.el7.noarch conflicts >> ovirt-engine < 4.2.6 >> --> Finished Dependency Resolution >> Error: Package: ovirt-engine-ui-extensions-1.0.6-1.el7.noarch (ovirt-4.3) >>              Requires: ovirt-engine-webadmin-portal >= 4.3 >>              Installed: >> ovirt-engine-webadmin-portal-4.2.1.7-1.el7.centos.noarch (@ovirt-4.2) >>                  ovirt-engine-webadmin-portal = 4.2.1.7-1.el7.centos >> Error: ovirt-engine-setup-plugin-ovirt-engine conflicts with >> ovirt-engine-4.2.1.7-1.el7.centos.noarch >>    You could try using --skip-broken to work around the problem >>    You could try running: rpm -Va --nofiles --nodigest >> >> >> >> Cheers, >> TK >> > > Just upgraded to 4.3.5. > > What do I do if I can't remove a host from a cluster?  Maintenance > window options Active and Maintenance are grayed out.  Host is currently > showing as Unassigned with an exclamation mark next to it.  Items listed > include: > > "Power Management is not configured for this Host. Enable Power Management" > "Gluster status is disconnected for this host. Restart Glusterd service" > "Host has no default route." > > > I try to modify some of this thinking maybe it will nudge the server out > of it's slumber but get this when trying to fix Power Management: > > > "Error while executing action: > mdskvm-p01.nix.mds.xyz: > Cannot edit Host. Host parameters cannot be modified while Host is > operational. > Please switch Host to Maintenance mode first." > > > Of course I can't put it into maintenance since it's grayed out. > > Anyway to remove this host and start fresh? > > Cheers, > TK > > >>> >>>     3) Create a Domain within the Cluster >>>     4) Add hosts to DC >>>     5) Import Gluster Volume >>> >>> >>> 4 & 5 not required if you change step 2 to import. >>> You are running into errors below because the engine sees the addition >>> of new hosts as gluster peer probing to a new cluster >>> >>> >>> >>>     Some issues.  I create a DC, add a Cluster then add the Host to it. >>>     Then >>>     I try to add another the second node but it says It's already in a >>>     cluster.  There is no other cluster and the second host is not listed >>>     anywhere. >>> >>>     So I try to remove the first host to res