[imp] High Availability / Load Balancing

[Dominic Ijichi]

Quoting Lee <lee@disinfo.com>:

> > 2) why bother replicating the sql database when you've still got a
> > single point
> > of failure on the imap server?  mailstore is notoriously difficult to
> > replicate
> > in realtime good enough to have a hot failover (if anyone does this
> > successfully
> > , PLEASE speak up and share!!).  so you put the sql db on the same
> > server/tier
> > as the mailstore and accept two single points of failure, both
> > catastrophic.
> >
> 
> We divide our users over two-box cyrus imap clusters. Each segment of
> our users resides on a primary box with a second hot spare waiting to
> take over in case of failure. Each box has a live copy of the data
> using drbd. The boxes then monitor each other using linux-ha's
> heartbeat. We havent had any problems with this setup thus far. We use
> a dual perdition cluster to make the individual segments appear to be
> one single server.
> 
> L

ooh, this looks nice, thanks!  i remember at one point people tried this with
linux md/swraid, but i don't think it was supported/recommended.  i currently
use solaris for my main server, linux for the backup, but wouldn't hesitate to
change to linux if this works well.  have you tried murder instead of perdition?

how do you set imp up for this?  do you look up per user and pick the imap
server from their record?  how does imp cope with failing over?

dom

------------------------------------------
This message was penned by the hand of Dom