Backups - WAL archiving. problem understanding documentation

From: "David" <luckychap(at)bluebottle(dot)com>
To: <pgsql-novice(at)postgresql(dot)org>
Subject: Backups - WAL archiving. problem understanding documentation
Date: 2005-07-06 02:58:33
Message-ID: 200507051029.j65ATh2m031788@bluebottle-fe1.bluebottle.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-novice

My database v8.3 is running well on Fedora C3, but now I have to instal a
backup system and
I am having trouble understanding the documentation. I am new to Postgresql
so please be
patient with my questions.

What is the difference between archive and backup? The terms seem to be
inter changed.

The "Archive" command "will always be placed in the postgresql.conf file (p
307). I could not see any sections or related commands there. Not even an
example. Will adding an archive command in this file actually cause wal
files to be copied? If so, when - when they fill or at a set interval or
when a full backup is done. Is there a way of forcing the Wal file to be
copied?

The following command (p308) uses -i and </dev/null. The -i allows interact
with the
terminal, the </dev/null suggests it can get input from null. It seems
contradictory to
me.

Cp -i %p mnt/server/archivedir/%f </dev/null

How can I ensure WAL archiving is enabled and working? I thought it was
automatic.

If I execute pg_start_backup, an archive command, and pg_stop_backup in a
script, will the
commands be carried out in sequence and do not overlap (p309).

Are backed up WAL files automatically deleted? Can a delete command be
added to the script?

If there is any more documentation or books for beginners that are explicit
on WAL backups I
will gratefully buy one.

Responses

Browse pgsql-novice by date

  From Date Subject
Next Message Reinhard Hnat 2005-07-06 11:34:10 Rules and transactions
Previous Message Mike Fahrenkrog 2005-07-05 22:42:53 Batch Scripting of DB Objects to Separate Files