Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi, looking at your kernel logs would be helpful, but this statement On 02/17/2012 10:19 AM, Lawrence Strydom wrote: > the DRBD device was mounted on both nodes but /proc/drbd output ... > this:* 0: cs:StandAlone ro:Primary/Unknown ds:UpToDate/DUnknown r----* screams "split brain". Long story short: You ran dual-primary without fencing or your fencing didn't work right. Furthermore, your monitoring (Nagios or whatever) has not been configured to detect DRBD hickups. Now you're as deep in trouble as it gets. Sorry. It's quite possible that the data on both your nodes has diverged during your disconnected operation. Salvaging will be difficult if you cannot bring OCFS2 up on one of your nodes. If you can afford to just shrug this off, it should be sufficient to simply resolve the split-brain: http://www.drbd.org/users-guide/s-resolve-split-brain.html You loose whatever was written to the victim peer while disconnected. If you cannot access its filesystem either way, that shouldn't pose much of an issue, I guess ;-S Advice before you go dual-primary again: - get your Pacemaker+Stonith right - devise monitoring of your DRBD status HTH, Felix