We have a 5 host DRS cluster running vsphere enterprise. We have just added a new host to the setup and it is currently outside the DRS cluster. As part of the testing I am experiencing a weird vmotion issue. I am able to vmotion machines onto the host as normal. When I vmotion a machine off the new host back onto one of our original hosts there is an outage of approximately 45 seconds and then the machine springs back to life.
I have had a look over the settings of the new host but I cannot see any differences to anything. Has anyone experienced any similar issues or can offer any advice?
Is this a complete outage (i.e. the VM is not accessible, not even the console) or is it a network outage? The 45 seconds you mention remind me of the Spanning Tree protocol. Can you confirm that all the uplink ports (for the target host) on the physical switches are configured for e.g. spanning-tree portfast?
Yes, I think it is a complete outage. During the outage the console window also breaks.
I think all of the ports are configure the same as the other hosts. I will have a look over them this evening and let you know. All of the other host machines are working fine so I will compare the port settings.