Open Stack

Hello All,

I'm running my test/dev environments in Mirantis openstack.
Fuel - 7.0
openvswitch 2.3.1-1~u14.04+mos2
3 controllers, 14 compute nodes.

Recently, many vms lose connectivity completely, and they cannot renew IPs
by doing a network restart too. Sometimes a compute node reboot fixes the
problem.
Sometimes, we keep restarting various neutron services like
openvswith-agent etc.
But all these, without understanding the problem.

Can someone guide me where to start troubleshooting?

A friend of mine suggested that this version of openvswitch has many bugs
and an upgrade might help. Is it even possible to do this on an online
cluster?

This issue has become very frequent and also spreads to more nodes. Some
pointer to understanding this issue will be of great help.

Thanks and Regards,
Raja.




-- 
:^)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20171015/007bf72f/attachment.html>