Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
2009/6/17 Kiran Kumar <mkirank1 at gmail.com> > Hi, > We are looking at using DRBD for our production servers , I have > been able to install and configure without any problems thanks to the > excellent resources available. > > I have been testing automatic recovery from split-brain in DRBD, I > have heartbeat running on multiple interfaces and am aware that > prevention of split-brain is recommended but I am looking at a worst > case scenario where a split brain occurs and I have to recover > automatically using "discard-least-changes". > > Suppose I have specified 4 resources r0 , r1 ,r2 and r3 is there any > hooks or any way of specifying in the configuration file that after > split brain sync from the node that has touched more blocks on one > resource say r1 and the other three resources just follow what r1 > decides . > This would be handy, but DRBD decides/resolves on resource level. So if you really need something like that, you will have to implement your handler script that would lock the others until it solves it. It can be a bit complicated, mainly if you consider all the possible aspects. If you will be able to manage it I am sure people on DRBD and HA mailing lists would appretiate if you would share it. > Another question I have, Is there any command to look at a particular > resource on each machine and determine which has more blocks written ? > In cat /proc/drbd you will find quite a lot, including number of blocks modified. Check DRBD manual pages. > I am attaching my drbd.conf . > > > Thanks, > Kiran > > _______________________________________________ > drbd-user mailing list > drbd-user at lists.linbit.com > http://lists.linbit.com/mailman/listinfo/drbd-user > > -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20090618/f1ffd991/attachment.htm>