Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
So, in both log files, you see: -- May 21 17:45:31 s6 kernel: block drbd0: helper command: /sbin/drbdadm before-resync-target minor-0 May 21 17:45:31 s6 kernel: block drbd0: helper command: /sbin/drbdadm before-resync-target minor-0 exit code 3 (0x300) May 21 17:45:31 s6 kernel: block drbd0: before-resync-target handler returned 3, dropping connection. -- This would be the reason that DRBD is disconnecting, so now to hunt down why (/sbin/drbdadm before-resync-target minor-0) returns exit code 3 :-). You might try defining a handler for before-resync-target, maybe /bin/true or something? You could also run this drbdadm command by hand on your other nodes and note the difference. This command should simply use your configuration to locate the correct handler and execute it. I assume if one is not defined, nothing bad should happen :-). In fact, I ran it on my test machine here to no ill effect. For some reason this operation fails on at least one of your six nodes. I don't know anything specific, but hopefully this gets you pointed in the right direction.