Andrew Sullivan ajs at crankycanuck.ca
Wed Sep 30 06:45:38 PDT 2009
On Tue, Sep 29, 2009 at 10:39:15PM -0700, roctaiwan wrote:
> 
> first, I don't think I had turn on WAL.

You can't actually turn WAL off.  Your remarks betoken several deep
misapprehensions about the way PostgreSQL works.  Without a basic
administrator's comprehension of the underlying database system, you
are never going to be successful at operating Slony.  On the contrary,
this is all going to be very painful for you (and, quite probably,
will be very unreliable) unless you get a really good grip on how
PostgreSQL works.

While Postgres doesn't demand that someone sit there twiddling the
knobs all the time just to keep it functioning reliably, it is
complicated enough that you need to understand it to operate it
successfully in production environments.  This is triply true of
Slony, which is a terribly complicated system with a great deal of
power and still a relatively primitive user interface.  If you don't
know how Postgres works and therefore can't understand what Slony is
doing, you're going to have a lousy experience of exactly the sort
you're currently enduring.  Therefore, I strongly suggest you get the
Postgres basics under control before trying to tackle replication with
Slony.

> second, how much space is big enough to handle 10M inserts? I have 6.4GB
> available right now. 

As someone already noted, 6G isn't a lot of space for many data sets.
The only real way to answer your question, however, is to calculate
what is in each row.  To do that, we'd need to see some sample data
and the database schema.

A

-- 
Andrew Sullivan
ajs at crankycanuck.ca


More information about the Slony1-general mailing list