[ENBD] Ammended setup: -> what would happen if...

[Liam Helmer]

Sorry, I was unclear -> the raid was from client to server, to deal with 
network disconnections and server reboots.

OK, cool. It looks like, since you're no longer doing the remote to 
local mirroring in enbd, I'll have to check into RAID-1 for the answers 
-> this is no longer an enbd question. Which should make answers more 
clear. I'd probably have to do a userspace synch: you're right, that I 
should be just looking at this as a hard disc. What I was most 
interested in is what behaviour the raid would exhibit... From this 
case, what I think is that the raid is going to take the most recently 
synched superblock and give it precedence... but won't integrate the two 
except with a manually issued command. Which is the best case scenario 
for me -> I can't deal with a suddenly corrupted disc very well, but I 
can create a userspace tool to perform a resynch of the raid.

Which makes me think that I've got this setup totally wrong for the 
application -> I need to be doing the mirroring from the filesystem 
layer or on top of it, rather than at the bottom level. So, maybe using 
coda, which has some automatic caching features, or just a simpler rsych 
type setup... where the cache is also on an encrypted volume. And, I 
need to get some way of effectively failing the disks without crashing 
the client. No really easy solution comes to mind though.

Thanks!
Liam