[ClusterLabs] start vs promote drbd m/s colocation constraint

[Brent Jensen]

On 2/2/2021 4:02 PM, Ken Gaillot wrote:
> I haven't used DRBD in a while, but I believe newer versions handle the
> equivalent of promotion internally, rather than via the resource agent
> and pacemaker. It could be that you simply don't need a promotable
> clone, just a regular clone. I don't know how that would actually work,
> but it's maybe something to investigate.

How's that? What would be the hooks? If someone has some experience as 
to why that would be great. It seems that people using pacemaker 2.x 
apparently successfully are using the 'start' order constraint instead 
of 'promote' and I have no clue WHY, except it seems to be working (it 
promotes the master correctly).

I have tested using/not using DRBD's:

  * fence-peer "/usr/lib/drbd/crm-fence-peer.9.sh";
  * after-resync-target "/usr/lib/drbd/crm-unfence-peer.9.sh"

And using "auto-promote yes;" A bit less clear on this.

These options didn't seem to make a working cluster when using the 
'promote' order constraint (e.g. fails on going standby).

Thanks, and still clueless,

Brent



-- 
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.clusterlabs.org/pipermail/users/attachments/20210202/87389de0/attachment.htm>