We do have an Openstack setup (Ocata Version) with 6 hypervisors connected to external Storage using cinder service cpu_allocation_ratio is 1:16 .In certain VM's at times we are facing issue with ssh to its floating IP. If we do a live migrate of that VM to another hypervisor we are able to ssh without any issue What can be the cause ?How can we trouble shoot this further?
The ask.openstack.org website will be read-only from now on. Please ask questions on the openstack-discuss mailing-list, stackoverflow.com for coding or serverfault.com for operations.