Hi, I haven’t looked at these filesystems so closely, but I want at around the same thing as you I guess: if the “main” imap-server fails, let another one take over the job (e.g. with IP takeover).
In my opinion maybe the NNTP support provides something in the future: it has the possibility to synchronise mail folders, so if it can do that with user's mailboxes (including flags) there is a good failover situation. Another option that might do the trick for small situations is use something like unison or rsync to synchronise the imap-directories from the "main-server" to the fallback server. If the first one fails the second one has a copy of the mail directories, however maybe outdated by e.g. an hour. NFS doesn't work well I think since when the fileserver goes down (hardware failure) the data is unreachable anyway. AFS was commercial I thought, and what CODA is concerned: I haven't looked at it so close, I thought it had only an option to cache a part of the data: that wouldn't be sufficient to let a second server "take over" de data for a while. Is this is a solution, I'd like to hear more about it. Paul -----Oorspronkelijk bericht----- Van: [EMAIL PROTECTED] [mailto:owner-info-cyrus@;lists.andrew.cmu.edu] Namens David Chait Verzonden: zaterdag 19 oktober 2002 8:24 Aan: [EMAIL PROTECTED] Onderwerp: Distributed File Systems Greetings, Has anyone here looked into or had experience with Distributed File Systems (AFS, NFS, CODA, etc) applied to mail partitions to allow for clusetering or fail over capability of Cyrus IMAP machines? I have seen docs for splitting the accounts between machines, however this doesn't seem like the best fault tollerant solution.
