Just curious if I missed a patch that changed the behavior or if I've created some anomaly.
We raised our cluster level EVC's a couple months ago across the board (3 different clusters). We set up batches/groups to begin power cycling of VM's, raising the hardware level and EVC in one go.
To prep, I ran a powerCLI script that would schedule the hardware level increase upon the next reboot.
This worked as expected in testing in our non-production cluster.
1. Run PCLI script against the group of VM's
2. Power Down the VM's
2.a Hardware level raised to maximum for ESXi 5.5 (10)
3. Power VM's on
3.a VM assumes the highest EVC that the HW level and Cluster permit
We performed step 1 on VM's several weeks ahead of time as it should have little effect. Raise the HW level does not usually hurt anything (found some bugs but not in scope of this thread)
What we observed was:
1. Run PCLI script against the group of VM's several days/weeks ahead of time
2. Some VM's reboot due to scheduled tasks
2.a Hardware level raises to 10
2.b EVC raises to maximum for HW level and Cluster setting
2.c No power off of the VM's is noted
3. Unexpected early profit
The third cluster has EVC disabled so we cannot use it for testing. Production and Non Production clusters are in different vcenters non-linked, both 5.5 U3.
We're happy to have the easiest upgrade ever but it was unexpected. Curious if others have seen this. For reference, I attached the PowerCLI script we used.