[DRBD-user] crm-fence-peer.sh & maintenance / reboots

Dirk Bonenkamp - ProActive dirk at proactive.nl
Wed Jul 25 12:30:21 CEST 2012

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

Op 24-7-2012 17:00, Florian Haas schreef:
> On Tue, Jul 24, 2012 at 2:17 PM, Dirk Bonenkamp - ProActive
> <dirk at proactive.nl> wrote:
>> Hi all,
>> I'm having trouble with the crm-fence-peer.sh script and maintenance /
>> reboots.
>> I have a simple Master / Slave set-up with 1 resource. Without the
>> crm-fence-peer.sh and crm-unfence-peer.sh scripts activated, things work
>> fine. When I reboot the master, the slave gets promoted, all good. When
>> I enable the scripts, the slave no longer gets promoted when the master
>> is rebooted or is put in standby mode. This error is in the logs:
>> CRIT: Refusing to be promoted to Primary without UpToDate data.
>> The error kind of makes sense, but before the reboot the data was
>> UpToDate/UpToDate.
>> The scripts work fine when I unplug the data connection between the
>> nodes. The location constraint is added when I unplug the cable and is
>> removed again when re-plugging the cable.
>> Am I missing something here?
> I'm not sure if you are, but we are. We're missing information about
> your DRBD and Pacemaker version, and your DRBD and Pacemaker
> configuration, specifically. Please collect and pastebin that, and
> follow up on this thread with the URL.

I guess I could have expected this question... sorry, here we go:

Ubuntu 10.04 running 2.6.32-41-generic / x86_64
DRBD 8.3.13 (self compiled)
Pacemaker 1.1.6 (from HA maintainers PPA)
Corosync 1.4.2 (from HA maintainers PPA)

Pastebin link with DRBD config, corosync / HA config:


Logfiles from both nodes when putting the master node (scan1) in standby

scan1 (master node): http://pastebin.com/YHz8sXCm
scan2: http://pastebin.com/jrq0BmGM



More information about the drbd-user mailing list