[Pgpool-general] some basic questions

[Jaume Sabater]

2009/2/4 Gerd König <koenig at transporeon.com>:

> It would be very nice if you could send your documentation to me
> directly, I'm pretty sure that I'll find somebody for translating it ;-)

I'll send you a URL as soon as I have it online. If I take too much
time, send me an email again. Give me a few days, though.

>> You may want to try online recovery with PostgreSQL's PITR.
>>
> Ahhhh, this means in addition to the replication I have to use PITR to
> copy the wal files to node2 ? Strange, I haven't thought of using two
> "techniques" at once ;-))
 think there are better ways ;-)

No. It means you want to use WAL archiving to archive logs, then
transfer them, then use PITR so that, when the node being recovered is
starting, it reads those WAL files and processes them to bring itself
back to speed with the master node.

I strongly suggest that you read the online documentation at
postgresql.org regarding WAL and PITR before getting started with
pgpool. It's not easy, but it's a must that you understand it. Still,
you don't really need to use PITR. It's an improvement IMHO, but not a
requirement.

-- 
Jaume Sabater
http://linuxsilo.net/

"Ubi sapientas ibi libertas"