On Fri, Oct 17, 2003 at 04:50:46PM +0100, Mike Williams wrote: > It all worked flawlessly for more than 6 months heavy use, shame the admin who > actually set it up fscked up the failover part. Can't actually remember what > happened to the primary, but I do remember that it took little more than a > reboot to fix.
Sounds like he didn't TEST it. Typical newbie "bug". If you set up failover and don't test it, you don't really have failover. Same thing with backups. See www.linux-ha.org for all sorts of Linux High-Availability projects. But remember, simple is good. Complex is bad. maildir over NFS with rsync to another machine that could take over manually is probably sufficient for most people's needs. If the boss wants better than that buy a real network storage device with super-availability and show via costs that to have that kind of uptime takes $. Managers/customers understand. -- Nate Duehr <[EMAIL PROTECTED]> -- [EMAIL PROTECTED] mailing list
