> Hi,
> 
> we have very bad experiences with hardware RAID-systems and therefore
> try to implement a high-available NFS service via drbd and pacemaker.
> 
> After reading lots of tutorials I am still not very comfortable with the
> whole setup. My first reaction is that it is quite fragile instead of
> high available.
> 
> I have several questions but the most important for me is:
> 
> What happens if the RAID storage on which drbd works fails on one node?

Data on the other node are ok. so no problem. Fix the raid / disk and start 
the replication.

> Will the system seamlessly migrate to the unaffected node?

Depends on your setup. But this is possible. Especially for a NFS server 
cluster.

> After fixing the RAID storage, reinitialization of it and a restart of
> the broken node, will the system recover automatically? Will drbd
> recognize that it has do replicate the whole data and perform this? Or
> is manual intervention necessary?

Yes. DRBD is even so smart that it will replicate only the missing data.

-- 
Dr. Michael Schwartzkopff
Guardinistr. 63
81375 München

Tel: (0163) 172 50 98

Attachment: signature.asc
Description: This is a digitally signed message part.

_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to