We are going to be backing up around 30 MS-SQL server databases via ADSL to a number of regional servers running CentOS (about 6 databases per backup server). 10 sites are 'live' as of now and this is how we have started...
The backups are between about 800MB and 5GB at the moment and are made as follows: 1) A stored procedure dumps the database to SiteName_DayofWeek.bak eg: SHR_Mon.bak 2) We create a local ZIP copy eg: !BSHR_Mon.zip. The !B means the file is EXCLUDED from backing up and is just kept as a local copy, cycled on a weekly basis. 3) We rename SHR_DayofWeek.bak to SiteName.bak 4) We split the .bak file into 200MB parts (.part1 .part2 etc.) and these are synced to the backup server via backuppc This gives us a generically-named daily backup that we sync (backupPC/rsyncd) up to the backup server nightly. We split the files so that if there is a comms glitch during the backing up of the large database file and we end up with a part backup, the next triggering of the backup doesn't have to start the large file again - only the missing/incomplete bits. Although the zip files are relatively small, we have found that their contents varies so much (bit-by-bit wise) on a weekly cycle basis that they take a long time to sync so we leave them as local copies only. Seems to work OK at the mo anyway! Any comments? Thanks ------------------------------------------------------------------------------ The Planet: dedicated and managed hosting, cloud storage, colocation Stay online with enterprise data centers and the best network in the business Choose flexible plans and management services without long-term contracts Personal 24x7 support from experience hosting pros just a phone call away. http://p.sf.net/sfu/theplanet-com _______________________________________________ BackupPC-users mailing list BackupPC-users@lists.sourceforge.net List: https://lists.sourceforge.net/lists/listinfo/backuppc-users Wiki: http://backuppc.wiki.sourceforge.net Project: http://backuppc.sourceforge.net/