Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Here's my drbd.conf resource "res1" { protocol C; startup { wfc-timeout 0; degr-wfc-timeout 120; } disk { on-io-error detach; } net { allow-two-primaries; } syncer { rate 15M; } on "machine1systems.com" { device /dev/drbd0; disk /dev/sdd; address 192.168.30.12:7788; meta-disk internal; } on "machine2.systems.com" { device /dev/drbd0; disk /dev/sde; address 192.168.30.14:7788; meta-disk internal; } } service drbd start drbdadm adjust res1(on both sides) drbdadm primary res1(on both sides) Then *on local primary side*, pvcreate /dev/drbd0 vgcreate drbdtest /dev/drbd0 lvcreate -n drbdlv1 -L 50M drbdtest Then,on *remote primary side*,I took snapshot lvcreate -S -n snap_drbdlv1 -L 50M /dev/drbdtest/drbdlv1 I am using 'drbd-8.3.0' on 'CentOS 2.6.18-92.el5' machines. *My problem* I tested above similar scenarios like above,I got split brain error at different intervals. like first time I got it after 10 minutes,so using http://www.drbd.org/users-guide/s-resolve-split-brain.html,I recovered from it. Then when I tested I got split brain after 30 minutes(though no operation was going on @ that time,I am closely watching /var/log/messages),then after recovery I got it after 1 hour and so on. So I am totally confused @ exact event of split brain.Though I am doing any IO,I got it sometimes. I am not using any shared file system at the moment.It might be required only if I want that level of concurrency. --------------------------------- Thanks and Regards, Himanshu Padmanabhi -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20090506/5c765da7/attachment.htm>