On Jan 3, 2017 7:37 AM, "TranceWorldLogic ." <tranceworldlogic@gmail.com> wrote:
Hi,

It look interesting.
Would you please point me some document that cover in detail ?

I was looking for mechanism where I can monitor guest application.
If application get failed or crash then I need to restart or start other VM.

Such monitoring should probably be done outside the VM. 

Is it possible in ovirt ?

There's a fence_rhevm script in the fence-agents pacakge, but it needs a bit of tweaking to work with 4.x
Y.

I though I will use hook but as you explain it clear me that my understand was wrong.
Thanks for your help.

Thanks,
~Rohit


On Mon, Jan 2, 2017 at 1:04 PM, Vinzenz Feenstra <vfeenstr@redhat.com> wrote:

> On Dec 30, 2016, at 11:03 AM, TranceWorldLogic . <tranceworldlogic@gmail.com> wrote:
>
> HI,

Hi there,

>
> I was exploring more about ovirt-gueste-engin.
> It look to me very easy to configure add add hook as script.
>
> But my doubt is, how to invoke those script from ovirt-engine ?
> Please some one help me to understand this part.
> I am looking into python SDK code to figure out same but still not got luck

Guest agent hooks aren’t triggered through the SDK, hooks are triggered when certain events happen on the hypervisor side.
e.g. A VM gets migrated from HOST A to HOST B or the VM gets suspended. In these cases VDSM _can_ send a message to the guest
agent asking it to process all hooks.

That those hooks are enabled are depending on the migration policy configured. Currently all but the ‘Legacy’  migration policies do cause the hooks to be executed, given a new enough guest agent, VDSM and cluster version.
HTH


>
> Thanks,
> ~Rohit
> _______________________________________________
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users



_______________________________________________
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users