[jira] [Commented] (CLOUDSTACK-2072) Virtual Routers Failover not working

["Koushik Das (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CLOUDSTACK-2072?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13721601#comment-13721601 ] 

Koushik Das commented on CLOUDSTACK-2072:
-----------------------------------------

I did a similar exercise as part of CLOUDSTACK-2427 on latest master but couldn't repro.

I used a XS 6.1 cluster with 2 hosts (say h1 and h2). Both hosts were newly installed.
- h1 put in maintenance - all VMs (system including router and user) migrated to h2 
- h2 put in maintenance - all VMs migrated back to h1 

If you see this issue on latest master again then please reopen with the full MS logs.
                
> Virtual Routers Failover not working
> ------------------------------------
>
>                 Key: CLOUDSTACK-2072
>                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-2072
>             Project: CloudStack
>          Issue Type: Bug
>      Security Level: Public(Anyone can view this level - this is the default.) 
>          Components: Management Server, Network Controller, Virtual Router
>    Affects Versions: 4.0.1, 4.1.0, 4.2.0
>            Reporter: Jeronimo
>            Assignee: Koushik Das
>             Fix For: 4.2.0
>
>
> I've experienced some erratic behavior when when setting a host in maintenance mode:
> I had about 6 Virtual routers running and 2 system vms in a given host ,  after putting this host in maintenance mode  the system vms failed over successfully.
> The problem came in when the VirtualRouters failed over , they did not start up at all, even after waiting 1 hour for HA to kick in .
> When i went on and start the virtual routers manually it would fail with this error:
> 2013-04-17 08:54:39,566 DEBUG [allocator.impl.FirstFitAllocator] (catalina-exec-2:null) Host name: kickseed, hostId: 1 is in avoid set, skipping this and trying other available hosts
> 2013-04-17 08:54:39,566 DEBUG [allocator.impl.FirstFitAllocator] (catalina-exec-2:null) Host Allocator returning 0 suitable hosts
> The work around i found is :
> Destroying the virtual router for network X
> Restart network X
> Is this the expected behaviour for virtual router fail over ?
> I'm using cloudstack 4.0.1 on Ubuntu .
> Thanks!

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira