I am about to start playing with Slony. For one, I have a stupid question:
where does one usually install Slony? On the master or on the slave?



--
http://mgogala.freehostia.com

.... on each node.

--
Greetings,

Robert

You are a fluke of the universe; you have no right to be here.

Mladen Gogala wrote:
From the documentation of the slon daemon
http://slony.info/documentation/slon.html
I would say that it needs to be installed on both.

You'll find people who know more on the Slony-I mailing list:
http://lists.slony.info/mailman/listinfo/slony1-general

Yours,
Laurenz Albe

On Thu, 12 Nov 2009 14:40:14 +0100, Laurenz Albe wrote:

Any more information? The information on those sites is horribly lacking
details. No details about the file names, no configuration files, the only
thing that they explain is about the slon process that controls the
replication and the slonik configuration processor. Nobody says anything
about what goes where and how. The largest single problem with Postgres is
the documentation.



--
http://mgogala.byethost5.com

Mladen Gogala wrote:
I have no knowledge of Slony-I, so I cannot provide much more information.
That's why I pointed you to the mailing list.

You probably read
http://slony.info/documentation/firstdb.html
but I'll post the link anyway.

I guess that you are frustrated by things that are probably hard
for anyone who is used to proprietary software:
- Different from Oracle where all tools and add-ons are integrated
and a single documentation covers them all, here you have to deal
with numerous different loosely related projects of varying
quality and sophistication.
- You'll frequently have to integrate the tools yourself by
writing glue, startup scripts and the like that suit your
own needs.
- Open source documentation sometimes lacks information about
details that the developers think will be better explained by
a look at the code or some playing around.

I am in no position to defend the quality of Slony's documentation,
but you should realize that there is a difference between core
PostgreSQL (whose documentation is excellent) and other projects.

One of the big benefits of open source is that you can influence
its development much better. One option is to whine at the appropriate
place (a mailing list that the developers read), and if you are
willing to improve the stuff yourself, you can do it.

Yours,
Laurenz Albe

On Fri, 13 Nov 2009 10:03:24 +0100, Laurenz Albe wrote:

Laurenz, I was able to set up Slony replication:

scott=> update emp set sal=4200 where empno=8005;
ERROR: Slony-I: Table emp is replicated and cannot be modified on a
subscriber node
scott=>

As far as your suggestion about doing things myself is concerned, I think
that I will write a tutorial or two early in the next year. I am eating,
breathing and sleeping PostgreSQL for 3 weeks now and am quite content
with the progress.





--
http://mgogala.byethost5.com