On Thu, May 06, 2010 at 08:50:21AM +0200, Michael Schwartzkopff wrote: > Am Donnerstag, 6. Mai 2010 01:22:13 schrieb Dimitri Maziuk: > > Cameron Smith wrote: > > > Yeah! > > > Since I already have that in place for http and mysql I just wanted to > > > know if there was anything unique I need to do for postfix config for > > > when it is running on primary (managed by heartbeat) and how do I handle > > > the sending of system emails on the secondary since postfix will be down > > > there? > > > > You don't really need drbd for mail storage: all you need's nfs-mounted > > maildirs. Nor should you need to shut down postfix on the secondary > > node. (If you want it to bind to cluster ip, you'll need to restart it > > on both nodes during failover, but that has nothing to do with sending > > system mail). > > > > Dima > > If you do not shut down postfix on one node might get problems with locks on > /var/spool/postfix which has to be mounted on any cluster node with postfix > running.
Thats not exactly correct ;-) keyword to go after on the postfix pages: alternate_config_directories You can easily have multiple instances of postfix on the same box. You may have a normal started-by-init per node "system" postfix for local system mail, using the "HA" postfix as next transport hop for everything that needs to be delivered there, spooling it if the "HA" postfix happens to be not available right now. Then you have the pacemaker-managed "HA" postfix, with its spool on drbd, delivering into mailboxes (of your preferred type) also on drbd. Caution: mailq, qshape and similar will only look at the default, you need to explicitly tell them if you want them to look at the alternates. We used to do that, but putting the mailserver into a VM and have only the VM managed by pacemaker turns out to be much simpler (especially simpler to explain to anyone who may need to operate/manage that thing). BTW, alert mails from the nodes about the cluster misbehaving should probably not try to be delivered into that HA postfix ;-) -- : Lars Ellenberg : LINBIT | Your Way to High Availability : DRBD/HA support and consulting http://www.linbit.com DRBD® and LINBIT® are registered trademarks of LINBIT, Austria. _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
