Shutdown policy "migration" / maintenance mode

[Mike Lupe]

First of all I would like to express my joy towards the elegance of the solution. Using the HA option and moving back the VMs to the shutdowned/rebooted node is a very simple but effective solution.

Is this the first step towards a "maintenance" mode for nodes? Could this mean maintenance mode would temporarly set all VMs on the respective node as HA and remove the HA flag after they came back up? I'm just curious.

 

[fiona]

First of all I would like to express my joy towards the elegance of the solution. Using the HA option and moving back the VMs to the shutdowned/rebooted node is a very simple but effective solution.

Is this the first step towards a "maintenance" mode for nodes? Could this mean maintenance mode would temporarly set all VMs on the respective node as HA and remove the HA flag after they came back up? I'm just curious.

Hi,
since there are some preconditions (see here) for HA, we can't simply set all VMs as HA managed. For example, a disk might be on a local storage. But if the preconditions are satisfied for all your VMs, why not just configure them to be HA managed (maybe using groups with a preferred node if you like)?

 

[Mike Lupe]

Hi Fabian,

thanks for the suggestions.

One suggestion would be to allow a setting (e.g. "Suspend and restart not migratable VMs? y/n"), where locally etc. storaged VMs would go into suspend mode and would automagically be continued as soon as the host's back up online.

I know, it's easy for me suggesting such things, implementation efforts are the other part

regards

 

[fiona]

You can open a feature request on the bugtracker. This makes it easier for us to follow the issue and increases the chance that this will be implemented (can't make guarantees though).