Open Stack

I think this is the way it works, ajay. When a compute node is rebooted, whatever state the application VMs were in, the same state is brought back (based on a setting in nova.conf). Typically all the services come back normally (unlike in your case), and they bring up the VMs in to running state. In your case, the service did not come up, hence the status of the VMs could not be reported by the nova-compute service running on the suspect compute node. SO, the last state (before reboot) is reflected

From: Ajay Kalambur (akalambu) [mailto:akalambu at cisco.com]
Sent: Friday, January 09, 2015 9:06 AM
To: Ajay Kalambur (akalambu); openstack at lists.openstack.org
Subject: Re: [Openstack] [nova]Compute node restart

Any comments

From: akalambu <akalambu at cisco.com<mailto:akalambu at cisco.com>>
Date: Thursday, January 8, 2015 at 4:18 PM
To: "openstack at lists.openstack.org<mailto:openstack at lists.openstack.org>" <openstack at lists.openstack.org<mailto:openstack at lists.openstack.org>>
Subject: [Openstack] [nova]Compute node restart

Hi
When a compute node is restarted I see that the controller reports the agents on these nodes as down I.e nova-compute and neutron-openvswitch-agent show down
However the application Vms on that compute node still show ACTIVE/RUNNING is this not supposed to be switched to SHUTOFF state?

Ajay

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20150109/a56e4a02/attachment.html>