[DRBD-user] Cluster filesystem question

Felix Frank ff at mpexnet.de
Fri Dec 2 09:40:46 CET 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,

On 12/01/2011 08:18 PM, Digimer wrote:
> When the primary node (the one with the floating IP) fails, all the
> cluster has to do is move the IP down to the backup node and it's ready
> to go.

I like the simplicity of this approach, it fits well into the chorus of
what I usually tend to do with DRBD. One question though:

Seeing as iSCSI is (supposed to be) a stateful protocol, is this
approach without risk? The way I see it, a failover can happen
mid-transaction for any given client.
I assume that such transactions will fail in some manner (which would be
OK).
But are there any other pitfalls I can't properly imagine?

Cheers,
Felix



More information about the drbd-user mailing list