Masakari with Kolla-ansible
Hello Team, I have deployed Openstack with Kolla-ansible on Centos8 and also enabled Masakari but I keep getting the error below in the masakari-hostmonitor.log on the controller nodes. Has anyone faced such a challenge ? 2021-08-02 18:08:52.409 7 WARNING masakarimonitors.hostmonitor.host_handler.handle_host [-] ' controller1.aos.rw' is unstable state on cluster.: oslo_concurrency.processutils.ProcessExecutionError: Unexpected error while running command. 2021-08-02 18:08:52.409 7 WARNING masakarimonitors.hostmonitor.host_handler.handle_host [-] hostmonitor skips monitoring hosts. 2021-08-02 18:09:57.476 7 WARNING masakarimonitors.hostmonitor.host_handler.handle_host [-] Corosync communication using 'ens33' is failed.: oslo_concurrency.processutils.ProcessExecutionError: Unexpected error while running command. Regards Tony Karera
On Mon, Aug 2, 2021 at 6:19 PM Karera Tony <tonykarera@gmail.com> wrote:
Hello Team,
Hello Tony,
I have deployed Openstack with Kolla-ansible on Centos8 and also enabled Masakari but I keep getting the error below in the masakari-hostmonitor.log on the controller nodes.
Have you enabled hacluster as well? It is a required component for hostmonitor unless you are providing your own hacluster (i.e., pacemaker+corosync cluster). Kind regards, -yoctozepto
Hello Radoslaw, I enabled it but it seems the corosync cluster is not working. I don't know if there is any other trick to try. Regards Tony Karera On Thu, Aug 5, 2021 at 3:26 PM Radosław Piliszek < radoslaw.piliszek@gmail.com> wrote:
On Mon, Aug 2, 2021 at 6:19 PM Karera Tony <tonykarera@gmail.com> wrote:
Hello Team,
Hello Tony,
I have deployed Openstack with Kolla-ansible on Centos8 and also enabled Masakari but I keep getting the error below in the masakari-hostmonitor.log on the controller nodes.
Have you enabled hacluster as well? It is a required component for hostmonitor unless you are providing your own hacluster (i.e., pacemaker+corosync cluster).
Kind regards,
-yoctozepto
On Fri, Aug 6, 2021 at 9:48 AM Karera Tony <tonykarera@gmail.com> wrote:
Hello Radoslaw,
I enabled it but it seems the corosync cluster is not working.
I don't know if there is any other trick to try.
There are no tricks. You might want to check the corosync status via: crm_mon -1 issued in the pacemaker container on the faulty controller host.
Regards
Tony Karera
On Thu, Aug 5, 2021 at 3:26 PM Radosław Piliszek <radoslaw.piliszek@gmail.com> wrote:
On Mon, Aug 2, 2021 at 6:19 PM Karera Tony <tonykarera@gmail.com> wrote:
Hello Team,
Hello Tony,
I have deployed Openstack with Kolla-ansible on Centos8 and also enabled Masakari but I keep getting the error below in the masakari-hostmonitor.log on the controller nodes.
Have you enabled hacluster as well? It is a required component for hostmonitor unless you are providing your own hacluster (i.e., pacemaker+corosync cluster).
Kind regards,
-yoctozepto
Dear Radosław, I got it working and also checked the corosync status and it was OK. Unfortunately when I test by killing the qemu-kvm service on the compute node. Only one instance on the compute node gets restarted and the rest remain shutdown. Regards Tony Karera On Fri, Aug 6, 2021 at 2:21 PM Radosław Piliszek < radoslaw.piliszek@gmail.com> wrote:
On Fri, Aug 6, 2021 at 9:48 AM Karera Tony <tonykarera@gmail.com> wrote:
Hello Radoslaw,
I enabled it but it seems the corosync cluster is not working.
I don't know if there is any other trick to try.
There are no tricks. You might want to check the corosync status via: crm_mon -1 issued in the pacemaker container on the faulty controller host.
Regards
Tony Karera
On Thu, Aug 5, 2021 at 3:26 PM Radosław Piliszek < radoslaw.piliszek@gmail.com> wrote:
On Mon, Aug 2, 2021 at 6:19 PM Karera Tony <tonykarera@gmail.com>
wrote:
Hello Team,
Hello Tony,
I have deployed Openstack with Kolla-ansible on Centos8 and also enabled Masakari but I keep getting the error below in the masakari-hostmonitor.log on the controller nodes.
Have you enabled hacluster as well? It is a required component for hostmonitor unless you are providing your own hacluster (i.e., pacemaker+corosync cluster).
Kind regards,
-yoctozepto
On Sun, Aug 8, 2021 at 12:37 PM Karera Tony <tonykarera@gmail.com> wrote:
Dear Radosław,
Dear Tony,
I got it working and also checked the corosync status and it was OK.
I am very glad!
Unfortunately when I test by killing the qemu-kvm service on the compute node.
There is one qemu-kvm per running instance - are you killing them all then?
Only one instance on the compute node gets restarted and the rest remain shutdown.
You would have to include both the Masakari instance monitor and Masakari engine logs. Do note the host monitor (so the Pacemaker stack as well) does not play any role here. -yoctozepto
Dear Radosław, Thanks a lot. I will check it out accordingly. I really appreciate the help Regards Tony Karera On Sun, Aug 8, 2021 at 6:38 PM Radosław Piliszek < radoslaw.piliszek@gmail.com> wrote:
On Sun, Aug 8, 2021 at 12:37 PM Karera Tony <tonykarera@gmail.com> wrote:
Dear Radosław,
Dear Tony,
I got it working and also checked the corosync status and it was OK.
I am very glad!
Unfortunately when I test by killing the qemu-kvm service on the compute node.
There is one qemu-kvm per running instance - are you killing them all then?
Only one instance on the compute node gets restarted and the rest remain shutdown.
You would have to include both the Masakari instance monitor and Masakari engine logs. Do note the host monitor (so the Pacemaker stack as well) does not play any role here.
-yoctozepto
participants (2)
-
Karera Tony
-
Radosław Piliszek