Hi,

I use DRBD 0.7.21-4 and HA 2.1.2 in order to have a master-slave-configuration 
on two nodes. We have a drbd master/slave resource and a resource group 
including an IPAddr.

The drbd master should move to the slave node if any service in the group 
fails and the group must run only on the drbd master

Currently, the master is not moved, if the group fails.

Andrew Beekow wrote: "in that case, add symmetrical=true to your colocation 
constraint"

I tried your suggestion adding the symmetrical=true value:

<rsc_order id="drbd_before_group" from="group" to="ms_drbd" action="start" 
to_action="promote"/>
<rsc_colocation id="group_on_drbd" from="group" to="ms_drbd" to_role="Master" 
symmetrical="true" score="INFINITY"/>

In result both systems could be started BUT both became slave!

cat /proc/drbd:
cs:Connected st:Secondary/Secondary ld:Consistent

crm_mon:
Master/Slave Set: ms_drbd
    drbd0:0     (heartbeat::ocf:drbd):  Started frigg
    drbd0:1     (heartbeat::ocf:drbd):  Started odin
Resource Group: group
    IPaddr0     (heartbeat::ocf:IPaddr):        Started frigg
    MailTo      (heartbeat::ocf:MailTo):        Started frigg

Without a master I cannot access to the DRBD ressources.

Secondly, how the group could be started? The group normally starts at the 
master -- and there is none...

Thanks for your help!
Klemens

-- 
Klemens Kittan
Systemadministrator

Uni-Potsdam, Inst. f. Informatik
August-Bebel-Str. 89
14482 Potsdam

Tel.    :   +49-331-977/3125
Fax.    :   +49-331-977/3122
eMail   : [EMAIL PROTECTED]

gpg --recv-keys --keyserver wwwkeys.de.pgp.net 6EA09333

Attachment: pgpwNsPLAVVva.pgp
Description: PGP signature

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to