Hi everybody,
This is my first entry here and I hope you can help me.
We have an ESX 5.1 infrastructure with 10 ESX servers distributed across two data centers, each one configured as a Cluster. For Administration we use the vSphere Client 5.1. There we have configured vSphere HA and vSphere DRS (partially automated). As data storage we have a Netapp Mode 9 in use. The virtual machines are located on different NFS volumes which are exported via the Netapp.
Last week, the stacked switch to which the Netapp is connected failed. The ESX servers could not reach the Netapp thus any more. After exchanging the broken switches for two separate switches, the ESX servers found the Netapp again and we saw, all virtual machines had been restarted. With the Netapp two of the ESX servers where connected to the broken switches. The virtual machines where all moved to the remaining ESX servers, so this was successful.
But after the switches has been swapped and the Cluster was complete again, a few of the virtual machines are constantly changing their ESX host. This behavior has not changed until today. The affected VMs are running properly but I would like to restore the previous state without the ever-changing ESX hosts.
I know that the version we use is not support anymore in any way. A migration is currently underway. But it is built in parallel.
What I've tried so far:
- Rebooted the vCenter Server
- Deactivated the vSphere DRS
- Activated DRS again
- Changed the DRS Threshold from 5 (agressive) to 4
- Changed the DRS Threshold to 1 (konservative)
I have no idea what else to do. Has anyone ever experienced such a behavior and still knows how to fix it?
Thank you for your answers
