First of all, thanks so much for the reply. To be clear, as stated in my
original post, if a user is interacting with the PostgreSQL database,
changes to the data folder continue even after the PG_START_BACKUP
command has been issued. This implies to me that the contents of any
copy of the data folder may be unreliable. i.e. the copy may not reflect
the state of the data folder either before the copy started or after the
copy has finished. It may reflect the state of the data folder in some
transient form. Assuming this is true, is the copy still usable for
restoration? If so, how does PostgreSQL get the data folder to a stable
state? Is it by the use of the WAL files that may be created during the
backup process and the restore.config file?
-----Original Message-----
From: Ian Lea [mailto:ian.lea@gmail.com]
Sent: Friday, April 08, 2011 3:55 AM
To: David Roland
Cc: pgsql-admin@postgresql.org
Subject: Re: [ADMIN] pg_start_backup - backups
> Attempting to run a backup with the database online. I first issue
> 'PG_START_BACKUP', then copy the contents of the postgresql data
> folder, then issue 'PG_STOP_BACKUP'. The database service remains
> running during the copy. If queries continue to hit the database after
> the PG_START_BACKUP command is issued changes to the data folder are
> possible, which can cause copy errors (for instance, a file not found
> exception). The file copy method first creates a string array of the
> data folder contents (all folders and
> files) and then processes the array, copying one file at a time. What
> am I doing wrong?
You need to use a program to do the copying that can accept changes to
files as it runs. rsync is a common choice and has the massive
advantage that it only takes changes and is therefore, after the first
run, quick.
See
http://www.postgresql.org/docs/9.0/static/continuous-archiving.html#BACK
UP-BASE-BACKUP
or use your favourite search engine to find more info.
--
Ian.