[DRBD-user] How to solve split brain?

Roberto Resoli roberto at resolutions.it
Mon Sep 5 13:54:21 CEST 2016

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


Il 05/09/2016 12:56, Michele Rossetti ha scritto:
> On node 1:
> root at mpve1:~# drbdsetup status vm-100-disk-1 --verbose --statistics
> vm-100-disk-1 node-id:0 role:Secondary suspended:no
>     write-ordering:flush
>   volume:0 minor:100 disk:Inconsistent
>       size:33554432 read:0 written:0 al-writes:0 bm-writes:0
> upper-pending:0
>       lower-pending:0 al-suspended:no blocked:no
>   mpve2 node-id:2 connection:StandAlone role:Unknown congested:no
>     volume:0 replication:Off peer-disk:DUnknown resync-suspended:no
>         received:0 sent:0 out-of-sync:3364 pending:0 unacked:0
>   mpve3 node-id:1 connection:StandAlone role:Unknown congested:no
>     volume:0 replication:Off peer-disk:DUnknown resync-suspended:no
>         received:0 sent:0 out-of-sync:3364 pending:0 unacked:0
> 
> On node 2:
> root at mpve2:~# drbdsetup status vm-100-disk-1 --verbose --statistics
> vm-100-disk-1 node-id:2 role:Secondary suspended:no
>     write-ordering:flush
>   volume:0 minor:100 disk:UpToDate
>       size:33554432 read:496 written:3820 al-writes:33 bm-writes:0
>       upper-pending:0 lower-pending:0 al-suspended:no blocked:no
>   mpve1 node-id:0 connection:StandAlone role:Unknown congested:no
>     volume:0 replication:Off peer-disk:Inconsistent resync-suspended:no
>          received:0 sent:0 out-of-sync:4928 pending:0 unacked:0
>   mpve3 node-id:1 connection:Connected role:Primary congested:no
>     volume:0 replication:Established peer-disk:UpToDate
>         resync-suspended:no
>         received:3820 sent:0 out-of-sync:0 pending:0 unacked:0

mpve3 ?

Resource is primary, i guess there is no split brain, but only:

1) lack of connection from mpve1 to the other two nodes:

Try to issue *on mpve1*:

drbdadm connect vm-100-disk-1

2) Lack of connection from mpve2 and mpve1 (node-id 0)

Try to issue *on mpve2*:

drbdsetup connect vm-100-disk-1 0

you can watch how the status is evolving after each command issuing:

watch -c drbdsetup status vm-100-disk-1 --color=always --verbose
--statistics

If the status keeps returning StandAlone, ther is probably a split
brain. You may try to add the --discard-my-data flag to "drbdsetup
connect" commands on *secondary* nodes to resolve that.

rob




More information about the drbd-user mailing list