Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
My /etc/drbd.conf file is standard:
global {usage-count no;}
common {
protocol C;
}
resource r0 {
on drbdtest1 {
device /dev/drbd0;
disk /dev/xendisks/drbd0;
address 192.168.10.101:7788;
meta-disk internal;
}
on drbdtest2 {
device /dev/drbd0;
disk /dev/xendisks/drbd0;
address 192.168.10.102:7788;
meta-disk internal;
}
}
After create-md on both nodes I see both in Connected Secondary/Secondary
state.
After "drbdadm -- --overwrite-data-of-peer primary r0" on one node, the
other node immediately drops the connection and enters Standalone state.
On that node I looked in /var/log/messages:
Jun 10 15:57:34 drbdtest1 kernel: drbd0: peer( Secondary -> Primary ) pdsk(
Inconsistent -> UpToDate )
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Writing meta data super block now.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Becoming sync target due to disk
states.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Writing the whole bitmap, full sync
required after drbd_sync_handshake.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Writing meta data super block now.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: writing of bitmap took 1 jiffies
Jun 10 15:57:34 drbdtest1 kernel: drbd0: 7168 MB (1834943 bits) marked
out-of-sync by on disk bit-map.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Writing meta data super block now.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: conn( Connected -> WFBitMapT )
Jun 10 15:57:34 drbdtest1 kernel: drbd0: conn( WFBitMapT -> WFSyncUUID )
Jun 10 15:57:34 drbdtest1 kernel: drbd0: helper command: /sbin/drbdadm
before-resync-target
Jun 10 15:57:34 drbdtest1 kernel: drbd0: before-resync-target handler
returned 3, dropping connection.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: peer( Primary -> Unknown ) conn(
WFSyncUUID -> Disconnecting ) pdsk( UpToDate -> DUnknown )
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Writing meta data super block now.
Jun 10 15:57:34 drbdtest1 kernel: drbd0: asender terminated
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Terminating asender thread
Jun 10 15:57:34 drbdtest1 kernel: drbd0: tl_clear()
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Connection closed
Jun 10 15:57:34 drbdtest1 kernel: drbd0: conn( Disconnecting -> StandAlone )
Jun 10 15:57:34 drbdtest1 kernel: drbd0: receiver terminated
Jun 10 15:57:34 drbdtest1 kernel: drbd0: Terminating receiver thread
It seems to me that a partial resync is the problem. Perhaps I can solve it
bij forcing a full synchronisation. I don't know how.
Piet Pijnacker
-----Oorspronkelijk bericht-----
Van: drbd-user-bounces at lists.linbit.com
[mailto:drbd-user-bounces at lists.linbit.com] Namens Johannes Thoma
Verzonden: dinsdag 10 juni 2008 15:32
Aan: drbd-user at lists.linbit.com
Onderwerp: Re: [DRBD-user] before-resync-target handler returned 3
On Tuesday 10 June 2008 01:35:01 pm P.Pijnacker wrote:
> After a succeeded experiment with Xen live migration, with drbd in dual
> primary mode, one node switched to WFConnection mode, the other in
> Standalone mode and I cannot fix it.
>
> In dmessg and /var/log/messages I see no 'split brain'.
>
> II did all the steps to to setup drbd again, but no effect.
>
> In /var/log/messages there is a line:
>
> before-resync-target handler returned 3, dropping connection.
>
> In the dbd-user archives or with Google I cannot find this message.
>
> What does it mean?
>
> How to restart the connection?
>
>
>
> Piet Pijnacker
Are you using the -c switch to drbdadm to use a non-standard config file? In
that case exactly that happens.
- Johannes
--
: DI. Johannes Thoma Tel +43-1-8178292-62 :
: LINBIT Information Technologies GmbH Fax +43-1-8178292-82 :
: Vivenotgasse 48, A-1120 Vienna/Europe http://www.linbit.com :
_______________________________________________
drbd-user mailing list
drbd-user at lists.linbit.com
http://lists.linbit.com/mailman/listinfo/drbd-user