[DRBD-user] DRBD running but....

Patrick Egloff pegloff at gmail.com
Fri Apr 29 18:41:55 CEST 2011

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


Hi all,

After my previous post asking if DRBD would be a solution for me, i finaly
decided and managed to start a 2 nodes active/passive cluster using DRBD and
Heartbeat 2.1.14.

After some days of hard work and support of a reader of this list (thanks
Olivier), everything seems to work OK.

But, in the machine boot process, i have now some errors showing up only on
the "master" node.

I try to translate the errors, as they appear in french. Just after the
quotas service, i have 3 lines :

rm : couldn't remove directory /var/run/heartbeat/ccm
rm : couldn't remove directory /var/run/heartbeat/ccm
rm : couldn't remove directory /var/run/heartbeat/rsctmp

I use the compatibility with version 1,using the parameter "crm no". I must
say that i first started heartbeat in version 2 mode... Could this be a
reason ?

What process does try the remove thes folders , and how can i correct this
??

Another question, when the DRBD link (eth1 for me) goes down for any reason,
and there are new datas being written on the primary node, what happens to
the secondary resource when the link comes back. Does DRBD synchonize these
datas on the secondary ?

Now i will try to implement "dopd" daemon and add "ipfail" in the heartbeat
configuration, for more security. Does this make sense ?

Thanks for any help !
Cheers,

-- 
Patrick Egloff - TK5EP
email : pegloff at gmail.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20110429/81fff629/attachment.htm>


More information about the drbd-user mailing list