Skip site navigation (1) Skip section navigation (2)

Synchronize unicode data using copy or pg_dump

From: Pascal Polleunus <ppo(at)beeznest(dot)net>
To: pgsql-general(at)postgresql(dot)org
Subject: Synchronize unicode data using copy or pg_dump
Date: 2004-04-23 14:28:16
Message-ID: 40892800.3030803@beeznest.net (view raw or flat)
Thread:
Lists: pgsql-general
Hi,

I need to synchronize some tables from a database (master) to another 
one (slave).
Both servers are running Debian Woody with PostgreSQL 7.2.1 (postgresql 
7.2.1-2woody4).
The databases are in unicode and doesn't contain any binary data.
The tables have primary/foreign key constraints, sequences and indexes, 
but no triggers/rules. There are OIDs but these are different on the 2 
databases. In fact they are not used by the application.

I need to do that ASAP, and hopefully only once (there are already some 
kind of home-made replication scripts, that are not the origin of my 
problem).
Later I would like to set up pgreplication or something... any advice on 
that is welcomed too ;-)


What would be the fastest way to achieve that?

On the master:
For each table, use COPY TO or pg_dump?

On the slave:
- delete everything from these tables
- drop the FK constraints
- (drop indexes?)
- vacuum (or later?)
- import from files (using COPY FROM or pg_restore)
- recreate the FK constraints
- update the sequences
- reindex (or recreate them)

Does that sound good?


Thanks for you help,
Pascal Polleunus



Responses

pgsql-general by date

Next:From: Bas ScheffersDate: 2004-04-23 14:53:03
Subject: Re: Insert Or update
Previous:From: Bruno Wolff IIIDate: 2004-04-23 14:02:57
Subject: Re: Question about inserts

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group