Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Fencing (stonith) would prevent split-brains from happening. In chapter
9 of Clusters from Scratch on the pacemaker website covers how to
configure fencing. Once setup, configure DRBD to use
'resource-and-stonith' and then use the 'crm-fence-peer.sh' fence
handler to hook DRBD into pacemaker's fencing.
This is the _only_ reliable way to avoid split-brains.
Cheers
On 06/25/2013 09:00 AM, Jonathan-Marc Lapointe wrote:
> Hi dear list members !
>
> I have a procluster project I'm working on right now and I have a little
> problem I'd like you to help me with.
>
> I'm configuring Pacemaker and Corosync.
>
> But when pacemaker starts DRBD, it detects a split brain and the status is:
>
> drbd driver loaded OK; device status:
> version: 8.3.13 (api:88/proto:86-96)
> srcversion: C0F510A918B92928FB51EE3
> m:res cs ro ds p mounted
> fstype
> 0:r0 StandAlone Primary/Unknown UpToDate/DUnknown r-----
>
> On both servers.
>
> But If I start it using the standard init script, it's all fine.
>
> Here are my configs:
>
> DRBD:
> ---------------------------------------------------------------------------------
> resource r0 {
> syncer {
> rate 1000M;
> }
> startup {
> wfc-timeout 15;
> degr-wfc-timeout 60;
> become-primary-on both;
> }
> net {
> allow-two-primaries;
> after-sb-0pri discard-zero-changes;
> after-sb-1pri discard-secondary;
> after-sb-2pri disconnect;
> cram-hmac-alg sha1;
> shared-secret "secret";
> }
> on web1dev {
> device /dev/drbd0;
> disk /dev/mapper/groupe1-data;
> address 192.168.77.212:7789;
> meta-disk internal;
> }
> on web2dev {
> device /dev/drbd0;
> disk /dev/mapper/groupe1-data;
> address 192.168.77.213:7789;
> meta-disk internal;
> }
> }
> ---------------------------------------------------------------------------------
>
> And Pacemaker:
>
> ---------------------------------------------------------------------------------
> primitive p_drbd ocf:linbit:drbd \
> params drbd_resource="r0" \
> op start interval="0" timeout="240s" \
> op promote interval="0" timeout="90s" \
> op demote interval="0" timeout="90s" \
> op notify interval="0" timeout="90s" \
> op stop interval="0" timeout="100s" \
> op monitor interval="20s" role="Slave" timeout="20s" \
> op monitor interval="10s" role="Master" timeout="20s"
> ms ms_drbd p_drbd \
> meta master-max="2" clone-max="2" notify="true" is-managed="true"
> ---------------------------------------------------------------------------------
>
> What could be possibly wrong ?
>
>
> CCDMD 20 ans !
>
> Jonathan-Marc Lapointe
> Technicien en informatique,
> Administrateur système,
> Administrateur réseau
>
> Centre collégial de développement de matériel didactique
> 6220, rue Sherbrooke Est, bureau 404
> Montréal (Québec) H1N 1C1
> 514) 864-8443
> www.ccdmd.qc.ca <http://www.ccdmd.qc.ca>
> jmlapointe at ccdmd.qc.ca <mailto:jmlapointe at ccdmd.qc.ca>
> Skype: jmlccdmd
>
>
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
>
--
Digimer
Papers and Projects: https://alteeve.ca/w/
What if the cure for cancer is trapped in the mind of a person without
access to education?