[DRBD-user] two node (master/slave) failover not working

Florian Haas florian at hastexo.com
Mon Oct 1 23:17:25 CEST 2012

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


On 10/01/2012 10:36 PM, Lonni J Friedman wrote:
> order FS0_drbd-after-FS0 inf: FS0_Clone:promote g_services

This your problem. That line is functionally equivalent to

order FS0_drbd-after-FS0 inf: FS0_Clone:promote g_services:promote

which means "you can start g_services at any time, but you need to wait
to promote it until FS0_Clone has also been promoted". Since the
"promote" operation is meaningless for g_services, and it's being
started prematurely, Pacemaker attempts to mount your filesystem before
DRBD is in the Primary role.

This needs to change to

order FS0_drbd-after-FS0 inf: FS0_Clone:promote g_services:start

... and you should be good to go.

You probably want to remove the cli-prefer-ClusterIP location constraint
as well.

And I find it very confusing the FS0 and FS0_Clone refer to a DRBD
master/slave set, whereas FS0_drbd is a filesystem. That seems backwards.

Hope this is useful.
Cheers,
Florian

-- 
Need help with High Availability?
http://www.hastexo.com/now



More information about the drbd-user mailing list