[rdo-list] Overcloud pacemaker services restart behavior causes downtime

Raoul Scarazzini rasca at redhat.com
Thu Aug 4 12:51:59 UTC 2016


Ok, so we are on mitaka. Here we have VIPs that are a (Optional)
dependency for haproxy, which is a (Mandatory) dependency for
openstack-core from which all the others (nova, neutron, cinder and so
on) depends.
This means that if you are rebooting a controller in which a VIP is
active you will NOT have a restart of openstack-core since haproxy will
not be restarted, because of the OPTIONAL constraint.
So the behavior you're describing is quite strange.
Maybe other components are in the game here. Can you open a bugzilla
with the exact steps you're using to reproduce the problem and share the
sosreports of your systems?

Thanks,

-- 
Raoul Scarazzini
rasca at redhat.com

On 04/08/2016 12:34, Pedro Sousa wrote:
> Hi,
> 
> I use mitaka from centos sig repos:
> 
> Centos 7.2
> centos-release-openstack-mitaka-1-3.el7.noarch
> pacemaker-cli-1.1.13-10.el7_2.2.x86_64
> pacemaker-1.1.13-10.el7_2.2.x86_64
> pacemaker-remote-1.1.13-10.el7_2.2.x86_64
> pacemaker-cluster-libs-1.1.13-10.el7_2.2.x86_64
> pacemaker-libs-1.1.13-10.el7_2.2.x86_64
> corosynclib-2.3.4-7.el7_2.3.x86_64
> corosync-2.3.4-7.el7_2.3.x86_64
> resource-agents-3.9.5-54.el7_2.10.x86_64
> 
> Let me know if you need more info.
> 
> Thanks
> 
> 
> 
> On Thu, Aug 4, 2016 at 11:21 AM, Raoul Scarazzini <rasca at redhat.com
> <mailto:rasca at redhat.com>> wrote:
> 
>     Hi,
>     can you please give us more information about the environment you are
>     using? Release, package versions and so on.
> 
>     --
>     Raoul Scarazzini
>     rasca at redhat.com <mailto:rasca at redhat.com>
> 
>     On 04/08/2016 11:34, Pedro Sousa wrote:
>     > Hi all,
>     >
>     > I have an overcloud with 3 controller nodes, everything is working fine,
>     > the problem is when I reboot one of the controllers. When the node comes
>     > online, all the services (nova-api, neutron-server) on the other nodes
>     > are also restarted, causing a couple of minutes of downtime until
>     > everything is recovered.
>     >
>     > In the example below I restarted controller2 and I see these messages on
>     > controller0. My question is if this is the expected behavior, because in
>     > my opinion it shouldn't happen.
>     >
>     > *Authorization Failed: Service Unavailable (HTTP 503)*
>     > *== Glance images ==*
>     > *Service Unavailable (HTTP 503)*
>     > *== Nova managed services ==*
>     > *No handlers could be found for logger
>     "keystoneauth.identity.generic.base"*
>     > *ERROR (ServiceUnavailable): Service Unavailable (HTTP 503)*
>     > *== Nova networks ==*
>     > *No handlers could be found for logger
>     "keystoneauth.identity.generic.base"*
>     > *ERROR (ServiceUnavailable): Service Unavailable (HTTP 503)*
>     > *== Nova instance flavors ==*
>     > *No handlers could be found for logger
>     "keystoneauth.identity.generic.base"*
>     > *ERROR (ServiceUnavailable): Service Unavailable (HTTP 503)*
>     > *== Nova instances ==*
>     > *No handlers could be found for logger
>     "keystoneauth.identity.generic.base"*
>     > *ERROR (ServiceUnavailable): Service Unavailable (HTTP 503)*
>     > *[root at overcloud-controller-0 ~]# openstack-status *
>     > *Broadcast message from
>     > systemd-journald at overcloud-controller-0.localdomain (Thu 2016-08-04
>     > 09:22:31 UTC):*
>     > *
>     > *
>     > *haproxy[2816]: proxy neutron has no server available!*
>     >
>     > Thanks,
>     > Pedro Sousa
>     >
>     >
>     >
>     >
>     > _______________________________________________
>     > rdo-list mailing list
>     > rdo-list at redhat.com <mailto:rdo-list at redhat.com>
>     > https://www.redhat.com/mailman/listinfo/rdo-list
>     >
>     > To unsubscribe: rdo-list-unsubscribe at redhat.com
>     <mailto:rdo-list-unsubscribe at redhat.com>
>     >
> 
> 




More information about the dev mailing list