[Nottingham] heartbeat, RAID 1 and ENBD

[Matthew Sackman]

Hi all,

Another blatent plug I'm afraid, but I guess some of you may be
interested in this...

I've just finished deploying a 2 node cluster using ENBD, software RAID
1 and heartbeat to ensure that a filesystem is always available so long
as only one node fails at a time. The system can cope with re-syncing
the RAID device to whichever machine didn't die, thus preventing loss of
data.

Code can easily be improved, all comments and suggestions welcome, if
you're at all interested then take a look at the write up and code at
wellquite.org/redundant/

Thanks,

Matthew

-- 

Matthew Sackman
Nottingham
England

BOFH Excuse Board:
Hard drive sleeping. Let it wake up on it's own...