Feature Request - Trigger VM Activity

[stemplar]

It would be nice have some more built-in/predefined commands as a result of an alarm.

As such I had following guest activities in mind

• - logoff
  - shutdown
  - restart
  - run command inside guest

In a test & dev VDI environment, there are and always will be VMs, where a service or an application goes ballistic for whatever reason. We can digest that for a while, but not too many of them. If we could automatically logoff/reboot a VM after a while, that would be nice. E.g. a logoff after 12 hours of CPU activity, a delayed logoff at 21:00, or a restart after 24 hours would be nice. I've not found much one Veeam One scripts. More seems to be available on the BR side.

[Shestakov]

Thank you for the feedback, Simon!

E.g. a logoff after 12 hours of CPU activity, a delayed logoff at 21:00, or a restart after 24 hours would be nice.

Do you want these actions to be done depending on particular performance values or not (For instance, if CPU usage>90% do logoff)?

We will think what actions we should make built-in taking your request into account.

You can run any possible action using scripts for now.

[Vitaliy S.]

It would be nice have some more built-in/predefined commands as a result of an alarm.

As such I had following guest activities in mind

- logoff
- shutdown
- restart
- run command inside guest

But you can already do that via script that can be triggered as a post alarm action. For example, using "psexec" you can run any command inside the Guest OS:

Code: Select all

C:\PSExec\PsExec.exe \\NAME_OF_THE_MACHINE -u USERNAME -p PASSWORD -i "ANY EXECUTABLE IN THE GUEST OS" 

[stemplar]

Well, with vCloud, we have one firewall per vApp. Therefore the execution of commands from 'outside' will fail, unless done through some VMware API. Our only option so far is to run things indirectly through the vmtools. Mentally, I'm getting closer to a self-written agent, that runs inside every vApp & VM. But, since Veeam One already has all the 'context' and has the features on the console tab, in my opinion, it would be easy for Veeam One to do those as a result of a trigger. Any custom script will need to perform all the login, navigation, and error handling again. And yes, the triggers would be mainly on performance values for an extended period of time.

[Vitaliy S.]

Yes, leveraging VIX API would allow to avoid setting up direct connection to the VM with a firewall enabled, however you will still need to provide user credentials to each Guest OS. Thanks for the feedback!