Running various business critical CentOS based Linux servers means good backups are a must. The ability to restore completely from backup to a DR site is a must.
To do this I keep a copy of the server backed up at root level on a locally hosted NAS.
A nightly rsync runs, with appropriate exlucsions that keeps a local copy of the system in sync. Data usage on the various Web/DNS/email servers I run works out to be around 50-200mb a night.
rm -f /var/log/sync/*
rsync -avzh –delete –stats server1.com:/ ย /store/backup/server1/ >/var/log/sync/server1 –exclude={/dev/*,/proc/*,/sys/*,/tmp/*,/run/*,/mnt/*,/media/*,/lost+found,/home/*/.gvfs,usage/*}
tail /var/log/sync/* | mail -s “Server sync report”logs@myemail.com
add a rsync line for each server.
the rsync logs are kept in /var/log/sync. To know if it was a success I can monitor the email. As I’m really only interested in success/fail i use tail to only show me the last few lines of each log.
To restore, do a minimal install of the same CentOS version, install rsync, and rsync directly over root from the backup. Reboot to bring the new system back online as per the last backup.