Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
I currently have drbd running on my clustered domU's. Everything works fine except when one hosts gets fenced or rebooted. When the fenced host comes back up I always get a Secondary/Unknown state. To get it back to Primary/Primary I have to umount the drbd0 disk from the active host then drbdadm down all and actually stop the service for drbd on both nodes and start the service again. The sync works perfect after this but how can I avoid restarting the services to get back to a known state? global { usage-count yes; } common { syncer { rate 100M; } } resource r0 { protocol C; handlers { pri-on-incon-degr "echo o > /proc/sysrq-trigger ; halt -f"; pri-lost-after-sb "echo o > /proc/sysrq-trigger ; halt -f"; local-io-error "echo o > /proc/sysrq-trigger ; halt -f"; } startup { degr-wfc-timeout 120; # 2 minutes. } net { after-sb-0pri discard-older-primary; cram-hmac-alg sha1; shared-secret "shared"; allow-two-primaries; } disk { on-io-error detach; } syncer { al-extents 257; } on samba-u01 { device /dev/drbd0; disk /dev/xvdb; address 192.168.7.2:7788; meta-disk internal; } on samba-u02 { device /dev/drbd0; disk /dev/xvdb; address 192.168.7.3:7788; meta-disk internal; } }