Josh wrote:
That's what I'm thinking might be the best method. Just need to find
some documentation on doing this for both MySQL and Postgres. One
thing to be careful of is that if the slave is lagging behind the
master and we have a failure, there will be some database changes
missing, which might be lost. Using drbd would mean that those
wouldn't be lost, but as you say possibly have some form of corruption.
PostgreSQL does not have built in replication. The 8.2.x distribution
comes with a perl script called dbmirror.pl that works along with a
trigger to replicate changes from one database server to the slave
whenever a change is made. You'll still have the same lag problem that
you'll have with a MySQL Master-Slave system; i.e. network transfer lag
between two machines. You might think about isolating the replication
traffic using the secondary Gbps interfaces on your servers and use a
separate VLAN on your switch or, better yet, use a separate switch. I
can't imagine that there would be much of a delay with two servers
sitting next to each other having a Gbps LAN to themselves.
I've also seen switches that will send data to multiple machines
simultaneously. I've seen firewalls that also do loadbalancing.
PFsense (a FreeBSD solution) comes to mind.
Its been my experience that a good database server can service several
mail servers. Most of the processing horsepower is required of the
server running the MTA and the POP3/IMAP servers. Most of that
horsepower gets used up decoding, decompressing, scanning,
recompressing, re-encoding mail messages for virus scanning and
anti-spam anyhow. If you're using TLS, you may also have to beef up the
processor.
I've separated the database server from the MTA server and the database
server's not even breathing hard.
Curtis
Curtis
_______________________________________________
DBmail mailing list
[email protected]
https://mailman.fastxs.nl/mailman/listinfo/dbmail
