From: | "Taras Kopets" <tkopets(at)gmail(dot)com> |
---|---|
To: | "Alban Hertroys" <alban(at)magproductions(dot)nl> |
Cc: | "Postgres General" <pgsql-general(at)postgresql(dot)org>, "S(dot)N(dot)E(dot) de Hartog" <simon(at)magproductions(dot)nl> |
Subject: | Re: Replicating changes |
Date: | 2006-10-27 15:01:12 |
Message-ID: | ce3f16fd0610270801o6216dd53t8aac4ce30075f3c2@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Hi!
I think you should try to use triggers with
dbi-link<http://pgfoundry.org/projects/dbi-link/>to achieve this.
AFAIK there is no such replication with other RDBMS as you need.
Taras Kopets
On 10/27/06, Alban Hertroys <alban(at)magproductions(dot)nl> wrote:
>
> Hello list,
>
> We're investigating a ways to replicate changes to the database to
> several "outside" systems.
>
> Some filtering will need to take place, as not all changes are allowed
> to go to all systems. Whether the system receiving the changes stores
> them in a database or not isn't particularly relevant to us. We're only
> interested in getting the data out in a specified format.
>
> The changes in the data that needs to be replicated can come in bursts
> of several thousands of records. Performance is important, but
> reliability as well. We need to get those changes to a number of 3rd
> parties, in a format that we're allowed to specify.
>
> So far we've looked into a few solutions:
> 1.) Adding triggers to the relevant tables that send a NOTIFY and store
> the changes in a local table. probably some meta-data will be required,
> like for example a time stamp.
>
> 2.) Use one of the existing replication systems. We're currently under
> the impression that (we've looked at Slony-I particularly) the slave
> system is supposed to be another (postgresql?) database. This wouldn't
> fit our needs, but maybe we're overlooking something?
>
> 3.) Somehow directly monitoring the WAL files, interpreting and
> detecting changes. So far we haven't found any documentation on how WAL
> files are stored, so we're not sure this is feasible.
>
> This is a call for advice, as we're surely not the first who are trying
> to accomplish this. If any clarification is needed, please ask. We want
> to get this right.
>
> --
> Alban Hertroys
> alban(at)magproductions(dot)nl
>
> magproductions b.v.
>
> T: ++31(0)534346874
> F: ++31(0)534346876
> M:
> I: www.magproductions.nl
> A: Postbus 416
> 7500 AK Enschede
>
> // Integrate Your World //
>
> ---------------------------(end of broadcast)---------------------------
> TIP 5: don't forget to increase your free space map settings
>
From | Date | Subject | |
---|---|---|---|
Next Message | Clodoaldo Pinto Neto | 2006-10-27 15:08:31 | plpython.so |
Previous Message | Richard Broersma Jr | 2006-10-27 14:55:32 | Re: pg_dumpall failing from possible corrupted shared memory |