Hi At,

> I got PGCluster (on 8.2.3) working in my environment.
> > A simple (but very busy - 8000 requests/sec) setup:
> Oh, Greate!
> It's my surprise.

We use this particular database for audit logging. I have 8 applications
servers (Java based) writing to this database.


> If you just need the shapshot DB for development purpose,
> you can create it with 'pg_dumpall' & 'pgrestore'.

That is the way I do it currently, but it takes too long to dump and
restore the database. I need something that can enable developers to
access the data quicker.

> BTW, I'm a bit confusing, since PGCluster doesn't replicate the "data
> reading query (i.e SELECT)", 
> 
> If there is no "data writing query(i.e. UPDATE/INSERT/DELETE)" in the
> developer's query,
> I recommend set 'true' to 'not_replicate_prepared_select' in the
> postgresql.conf.
> Would you explain more detail about contents of each queries. 
> 

Developers would not normally update cluster_2.

The cluster_1 (production side) does 90% insert queries which needs to
be replicated to cluster_2, which is synchronous, therefore when
developer queries cluster_2
(ex: "select * from table_name order by field3", "table_name" has +-
4,000,000 rows but there is no index on field3),
this query will cause heavy disk i/o on cluster_2 and therefore cause it
to be less responsive, which in turn causes cluster_1 to wait longer for
a response from cluster_2.


Thanks.


To read FirstRand Bank's Disclaimer for this email click on the following 
address or copy into your Internet browser: 
https://www.fnb.co.za/disclaimer.html 

If you are unable to access the Disclaimer, send a blank e-mail to
[EMAIL PROTECTED] and we will send you a copy of the Disclaimer.
_______________________________________________
Pgcluster-general mailing list
[email protected]
http://pgfoundry.org/mailman/listinfo/pgcluster-general

Reply via email to