[DRBD-user] Concurrent state changes detected and aborted

Dominik Klein dk at in-telegence.net
Thu Jun 4 09:38:55 CEST 2009

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


Hi

I'd really appreciate an answer on what happened in this case.

A guess from my side is that both nodes happened to try the state change
in the exact same moment (microsecond?) and that's why it happened and
it can actually not be avoided.

Does that make sense?

Regards
Dominik

Dominik Klein wrote:
> Dominik Klein wrote:
>> As requested on IRC, here's "crm configure show".
> 
> I agree it doesn't look too good that all the RAs have provider "dk" but
> I can assure you that all the RAs are completely identical to the ones
> from the heartbeat/pacemaker repository except for calling a script
> sending a text message to a mobile phone and an email on stop/start ops.
> This script does NOT alter any return code in the RA. It's like
> 
>     stop)
>         VirtualDomain_Stop
>         ret=$?
>         /root/bin/cluster-action $OCF_RESOURCE_INSTANCE $1
> dk at in-telegence.net &
>         /root/bin/cluster-action $OCF_RESOURCE_INSTANCE $1
> some_phone_number &
>         exit $ret
> ;;
> 
> I do this to a) receive those messages and b) to not have something
> overwritten during software updates and forget to update my change to it :)
> 
> Thanks for your help.
> 
> Regards
> Dominik
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
> 




More information about the drbd-user mailing list