[ClusterLabs] Antw: Re: pacemaker remote node ofgline after reboot

[Klaus Wenninger]

On 05/15/2017 01:46 PM, Ulrich Windl wrote:
>>>> Klaus Wenninger <kwenning at redhat.com> schrieb am 15.05.2017 um 09:12 in
> Nachricht <a7b2d0d3-1bbc-8154-fa85-104407ceddc1 at redhat.com>:
>
> [...]
>> Did you set the cluster-recheck-interval to a reasonable short value
>> (needed for connect-failures to timeout)?
> How short is "reasonable"? ;-)

Basically burns down to the question in that case:
Did you wait as long as recheck-interval takes to expire? ;-)

>
> We set the interval to somewhat longer, because it gives the administrartor the change to cleanup after mistakes before the cluster does ;-)
> Meaning: When the admin does it, it works mostly without a reboot; if the cluster does it, it will fence the node most of the time...
>
> Regards,
> Ulrich
>
>


-- 
Klaus Wenninger

Senior Software Engineer, EMEA ENG Openstack Infrastructure

Red Hat

kwenning at redhat.com