Hello,

Here is the script I made for a 24x24 running service with several medium size database

Features:

• keeps a logfile
• backups multiple databases
• does a vacuumdb before backing up
• doesn't stop the openerp server and
• removes old files (more than 10 days).

Simples and efective...

The script is executed via cron with

 00 01 * * * nice -19 /home/mbmaster/scripts/backupdb > /dev/null 2>&1

the "nice -19" is important to lower the priority of the backup

 # Backup script starts here.

 #!/bin/bash
 # Location of the backup logfile.
 logfile="/home/mbmaster/backups/logfile.log"

 # Location to place backups.
 backup_dir="/home/mbmaster/backups"

 touch $logfile
  timeslot=`date +%d%m%y%H%M%S`
  databases=`psql -U postgres -q -c "\l" | awk '{ print $1}' | grep -vE '^\||^-|^List|^Name|template[0|1]|^\('`

  for i in $databases; do
    timeinfo=`date '+%T %x'`
    echo "Backup and Vacuum started at $timeinfo for time slot $timeslot on database: $i " >>
    $logfile
    /usr/bin/vacuumdb -z -U postgres $i >/dev/null 2>&1
    /usr/bin/pg_dump $i -U postgres | gzip > "$backup_dir/openerp-$i-$timeslot-database.gz"
    timeinfo=`date '+%T %x'`
    echo "Backup and Vacuum complete at $timeinfo for time slot $timeslot on database: $i " >> $logfile
 done

 #-------------------------------------------------

 # delete files more than 10 days old
 find $backup_dir/openerp* -mtime +10 -exec rm {} \;

Remember to test your backups otherwise they aren't backups.

Hi,

There is module available "auto_backup" which takes backup of your database regular basis.

Email : info@acespritech.com<br&gt; Skype : acespritech<br> Blog : acespritechblog.wordpress.com

Another good option is barman, see this nice presentation about it from the 2014 OpenDays

• Hot, full, differential and incremental backups
• Support multiple databases
• Remote management
• And a lot of other features. Again, see the presentation.

I'd like to share my solution with a community.

Key features:

• It uses python-rotate-backups tool to apply more complex strategy in deleting old backups. E.g. you will have 7 backups for last 7 days, 4 backups for each week of last 30 days, 12 backups for each month of last year etc.
• It can backup posgtresql dump only  as well as full dump (with a filestore)
• it reads odoo configuration file to get data_dir and parameters to access database

To use script install dependencies:

pip install rotate-backups

Then download script to /usr/local/bin and make it executable:

cd /usr/local/bin/

wget -q https://gist.githubusercontent.com/2abdd91d00dddc4e4fa4/raw/odoo-backup.py -O odoo-backup.py

chmod +x odoo-backup.py

Then add lines at /etc/crontab like these:

# m h dom mon dow user    command
11 6    * * *    ODOO_USER odoo-backup.py -d DATABASE_NAME -p /PATH/TO/BACKUPS -c /PATH/TO/ODOO.conf --no-save-filestore --daily 8 --weekly 0 --monthly 0 --yearly 0
12 4    * * 7    ODOO_USER odoo-backup.py -d DATABASE_NAME -p /PATH/TO/BACKUPS -c /PATH/TO/ODOO.conf 

To test script execute something like this:

sudo su - ODOO_USER -s /bin/bash -c "odoo-backup.py -d DATABASE_NAME -p /PATH/TO/BACKUPS -c /PATH/TO/ODOO.conf ​"

If you have a lot of products with images you can check out that question about moving images from database to filestore: https://www.odoo.com/forum/help-1/question/how-can-i-add-product-images-from-filestore-1238

If your OpenERP server is a Linux server, then here is a link to some beta bash scripts that may accomplish the backup portion.

https://github.com/nagalman/oerp-admin.sh

Please note these scripts are released under the GNU GPL. If you improve them, please share the changes with me and others.

As far as doing a hot backup, a true "hot" backup is not possible with atomized databases (like PostgreSQL). You can dump the database to a dump file, then backup the dump file. While Postgre is dumping the database, no operations can be performed, so its best to do a database dump when OpenERP is not being used (ie 4am Sunday morning). If you set up a slave PostgreSQL server, then the slave server may be able to be dumped, but I do not know this for certain. The dump operation takes only a few seconds (of course this depends on the size of your database and the speed of your server, etc).

If you have a master slave(s) setup, you may be able to dump a slave database (slave databases just replicate what is in the master to be able to support fast read operations. Only the master can do add, delete, update) while the master is able devote resources to add, delete, update.

The scripts above are only set up to backup a single or master PostgreSQL database. I haven't done any testing in an environment that has master slave(s).

I haven't experimented with incremental backups of a database dump files. I suppose backup tools like TAR and others could perform this since the database dump file is just a file.

I hope this helps!

If your system Ubuntu, you install module autopostgresqlbackup:

sudo apt-get update

sudo apt-get install autopostgresqlbackup

This is the best tools for me. This create, auto and hot: latest, daily, weekly and monthly backup with rotation

1 - Backup-Script /var/lib/postgresql/postgres_db_backup.sh

#!/bin/bash

# Location to place backups.
backup_dir="/var/backups/postgres_db/"

#String to append to the name of the backup files
backup_date=`date +%Y-%m-%d_%H-%M`

#Numbers of days you want to keep copie of your databases
number_of_days=3
databases=`psql -l -t | cut -d'|' -f1 | sed -e 's/ //g' -e '/^$/d'`
for i in $databases; do
  if [ "$i" != "template0" ] && [ "$i" != "template1" ] && [ "$i" != "postgres" ]; then
    backupfile=$backup_dir$i.$backup_date.sql.gz
    echo Dumping $i to $backupfile
    pg_dump $i|gzip > $backupfile
  fi
done
find $backup_dir -type f -prune -mtime +$number_of_days -exec rm -f {} \;

2 - open your terminal

• su

• su postgres

• crontab -e

add

• 45 */4 * * * /var/lib/postgresql/postgres_db_backup.sh

I have tested a lot of tools, and now I use PostgreSQL-Backup http://postgresql-backup.com/postgresql-blog/backup-tool It's a simple tool which creates remote PostgreSQL database backups,Zips, Encrypts and sends backups to a folder, FTP, Dropbox, Box, Google Drive, MS OneDrive, Amazon S3 or Windows Azure Storage. Runs on a flexible Schedule and sends email confirmations on job success or failure. Also, backups file folders, allows to view the results on the web and more.

Hi there,

version 1.4.0 of Barman, Backup and Recovery Manager for PostgreSQL, includes support for transparent file-level incremental backup, which can bring significant reductions in terms of disk space, backup time and network consumption.

Please look at the following resources for more information:

• http://www.pgbarman.org/barman-1-4-0-released/
• http://blog.2ndquadrant.com/incremental-backup-barman-1-4-0/ (a blog article I have written about this feature)
• http://www.slideshare.net/openobject/odoo-disaster-recovery-with-barman (slides I presented at Odoo conference 2014 in Brussels)
• https://www.youtube.com/watch?v=Ka-4R43XJFs (introductory video on Barman)

One important aspect of Barman is that it also implements declarative retention policies (by REDUNDANCY or RECOVERY WINDOW), as well as other features such as WAL compression. Most importantly it allows to recover at any point in time, with just a simple command.

Thank you,

Gabriele

I use Backupninja: http://manpages.ubuntu.com/manpages/gutsy/man1/backupninja.1.html

Simple,and works fine together with Duplicity (https://help.ubuntu.com/community/DuplicityBackupHowto) to provide offsite backup

1. here is simple backup script

#!/bin/bash

cd /home/prasoon/p                         # give the path where you want to save your backup zip file

date=date+"%Y%m%d_%H%M%N"  # changing its date format

pg_dump admin | gzip > prasoon.gz    # using the dump command that convert the backup into zip format

save this file as backup.sh inside the /etc/cron.daily

and using cron tab you will be able to take backup

2. After creating your script

use

crontab -e

1 * * * * /etc/cron.dally/backup.sh

Here is a better solution, you can set a sftp extra location. I also added the local to DropBox folder, now I have 3 backups - https://github.com/Yenthe666/auto_backup

We also solved this over a nightly chron job, who is backing up the database every night and replaces older db copys.

You can insert a cron line where you can use pg_dump

Hey,

I have used the official answer but when i test this is the error i get, please help

sudo sh /var/scripts/dump_db.sh
Stopping openerp-server: openerp-server.
pg_dump: [archiver (db)] connection to database "SPIDDAFRICA" failed: could not connect to server: No such file or directory
        Is the server running locally and accepting
        connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5433"?
gzip: /var/pgdump/odoo_SPIDDAFRICA_20150219_0658056726971.sql: No such file or directory
pg_dump: [archiver (db)] connection to database "EMBARQ" failed: could not connect to server: No such file or directory
        Is the server running locally and accepting
        connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5433"?
gzip: /var/pgdump/odoo_EMBARQ_20150219_0658104368481.sql: No such file or directory
pg_dump: [archiver (db)] connection to database "Opencare" failed: could not connect to server: No such file or directory
        Is the server running locally and accepting
        connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5433"?
gzip: /var/pgdump/odoo_Opencare_20150219_0658180975532.sql: No such file or directory
Starting openerp-server: openerp-server.

Hey,

You can just install the auto_backup module from the extra-addons. This way you can configure your backups from inside openerp.

No need to stop your openerp-server running. Everything just keeps working while making a backup.

if you have some extra bells and whistles needed, you can still get a script or something