[Linux-HA] Delaying HA V2 resources not to start up at the same time

[Andrew Beekhof]

On Wed, Jul 2, 2008 at 09:47, Ivan <hunvagyok at freemail.hu> wrote:
> Hi,
>
> I haven several heavily loaded Xen resources in a 2 node V2 cluster,
> they all depend on the same evms resource.
>
> I would like to implement some kind of delaying to save the hard time
> from the dom0 host starting them up at the same time.
>
> There's a start_delay and stop_delay parameters. I assume that because
> they all depend on the same resource (evms) the counter would start for
> all at the same time so it should be like this:
>
> vm1 start_delay=0
> vm2 start_delay=20
> vm3 start_delay=30
> vm4 start_delay=40
>
> vm1 stop_delay=40
> vm2 stop_delay=30
> vm3 stop_delay=10
> vm4 stop_delay=0
>
> And also I assume that HA adds these delay values to timeout values.
> What I mean is that if my vm4 has 60sec stop timeout and have 40 stop
> delay it will not consider the resource failed after 20 seconds right?
>
> Is that correct?

i dont know anything about stop_delay, but if you specify the
start_delay for the operation

  ie. <op name=start interval=0 timeout=60s start_delay=40s/> , and/or
       <op name=stop interval=0 timeout=60s start_delay=40s/>

then everything will work as you expect (the 40s is automagically
added to the 60s timeout)

> Would these delays be considered by HA when it comes to migration?

No, you need to specify it for every operation you want to delay the
execution of.