Dumpy is a Python database backup script that uses configuration files to specify databases to backup and options. Backup scripts are classes which define a backup method.
Each BackupBase subclass returns a NamedTemporaryFile object. It's up to any post processors to use this object in any way (e.g. copy it to another location on the file system).
Post processors can be chained and all take the form:
If the post process doesn't alter the file passed in it should return it unchanged.
Example configuration file
The following is an idea of what the configuration file, located at ~/.dumpy.conf might look like. This is very likely to change:
[database db1] type = mysql name = dbname1 user = db1 password = db1 postprocessing = TimestampRename, Bzip, SystemFileCopy, S3Copy [database db2] type = postgresql name = dbname2 user = db2 password = db2 postprocessing = TimestampRename, Bzip, SystemFileCopy [mysqldump options] path = /opt/local/lib/mysql5/bin/mysqldump flags = -Q --opt --compact [pg_dump options] path = /opt/local/lib/postgresql83/bin/pg_dump [TimestampRename options] format = %Y%m%d [Bzip options] path = /usr/bin/bzip2 [S3Copy options] access_key = access_key secret_key = secret_key bucket = bucket
Very alpha. Looking for other coders to help and flesh out ideas.
I've written my last database dump and backup script that I want to. My hope is that this will be a general and feature rich backup script that's easily extendable and will work across multiple databases and backup schemes.
- Support some file based backups with auto rotation.
- Finish S3 Backup and flesh out S3 options.
- Lots more to think of.