Morning All,
This has been bugging me for some time, and the time has come to ask for assistance... Common scenario. Lots of ESX 5.5 hosts to update automagically. vCenter Update Manager is installed. I run 2 pinned virtual appliances on each host for platform applications, let's call them A and B, and each VA requires to remain pinned to the host. When I put a host into maintenance mode, the 'other' virtual machines on each host migrate off, leaving these 2 VA's. VA 'A' that requires to go down last is waiting for VA 'B' to go down, but 'B' doesn't go down. VA 'A' WILL shut down when it see's it's the last entity on the host, but because 'B' won't shut down, it's left in a state of limbo. It's a missing feature in the product and a product enhancement request is in place but...
The fact that VA 'B' won't shut down means I'm unable to automate this process, usually it would be a case of setting VCUM in motion and letting it do it's thing across the hosts, but this niggle is causing a major issue in this process.
Is there any way, via perhaps CLI, that this requirement can be included in the update process? Something sent to VA 'B' on each host as VCUM moves across the hosts to shut it down, instead of having make this whole update process manual?
I'm struggling to visualise what this might look like, whether the whole process becomes a CLI script or if it can be added/injected in some manner to the process.
If I've failed to explain the situation properly, feel free to ask, but I'm banging my head against the wall on this. Any guidance appreciated.