On Apr 5, 2011, at 1:01 AM, per...@pluto.rain.com wrote: > Chuck Swiger <cswi...@mac.com> wrote: > >> It's fairly common to scale up a mail infrastructure from one box >> handling both SMTP and IMAP (or POP) to a SMTP-only box writing to >> NFS-mounted user mailboxes, and have one or more dedicated reader >> boxes which only run IMAP/POP daemons which access that same NFS >> filesystem holding the user mailboxes. > > Yikes! The _proper_ way to scale up from one box to multiple goes > something like this: [ ... ] > mail.<domain> becomes a multiplexer, so that the outside world can > send to <user>@mail.<domain> (or simply <user>@<domain> if you > prefer) without needing to know about the internal structure. The > multiplexing can be very simple, and thus very fast, by doing a > table-lookup based on the first character of <user>. There's no > need for NFS anywhere in the setup.
The problems with this are that each individual reader box has no failover capabilities if something goes down (although you can have standbys and restore from backups), and the convention for allocating the pool of users onto boxes is relatively static, which means you don't distribute clients evenly onto the servers to even balance load as that changes dynamically over time. Growing a mail infrastructure from one to two to many boxes is complex. People tend to take advantage of the resources they have; if you have an EMC or NetApp filer handy, it's might well be reasonable to use it, especially if you can run with a mailbox format that doesn't require NFS locking. Regards, -- -Chuck _______________________________________________ freebsd-net@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-net To unsubscribe, send any mail to "freebsd-net-unsubscr...@freebsd.org"
