Re: Clustered PostgreSQL

Jojo Paderes wrote:

>We're currently using Ubuntu Breezy Hedgehog for our existing
>PostgreSQL database server (8.0.x). A Java-based applications uses the
>database server via JDBC.
>
>I'm not sure if this question is within the context but I would like
>to know if clustering can provide a hot-standby instance for a
>production-running PostgreSQL server? Say if the main db server goes
>down, another hot-standy with mirrored data from the main server can
>be used to replaced the defective db server.
>
>
PgPool or Slony-I could be used for this.  The difference being that
PgPool is statement level and thus may not be usable in some
environments.  It is also synchronous.  Slony-I is row-level and async,
but if your system goes down you may lose a small amount of unreplicated
data during the switchover.

YMMV...

HTH,
Chris Travers
Metatron Technology Consulting

>On 3/14/06, Louis Gonzales <louis.gonzales@linuxlouis.net> wrote:
>
>
>>Slony-I is another cluster software for postgreSQL.  What OS are you
>>running?  I just deployed it on Solaris 9, between a Sun Ultra
>>Enterprise E450 and a Sun Ultra 30.
>>
>>It's really great.  Currently, it only supports Single
>>Master-to-multipleSlaves.  Meaning, the single master is the only node
>>where database updates can occur, the changes are then propagated to the
>>slave nodes.
>>
>>http://gborg.postgresql.org/project/slony1/projdisplay.php
>>
>>
>>
>
>
>--
>http://jojopaderes.multiply.com
>http://jojopaderes.wordpress.com
>
>---------------------------(end of broadcast)---------------------------
>TIP 2: Don't 'kill -9' the postmaster
>
>