cross backups
Since I'm about to update my server to new hardware, I'm being reminded of the importance of backups. I'm doing a really brute force backup of the userspace files by doing an
scp -pr directorytree desktopmachine:backuppartition
I realized that this is taking up a bunch of extra space because it just decends the symlinks like they were in the directory tree.
Is there something easy, simple and convenient for backing up from one machine on a LAN to another? If my house burns down and both machines are cooked, I'm hosed, but short of offsite backups, this seems to be the most expedient way of backing up, if I set it up as a chron job. I can save my desktop data on my server, and my server on my desktop.
I've got a tremndous number of duplicate files, due to brute force backups. At the price of harddrive space, it's not worth the time to go through and cull duplicate files. I'd love something that would keep the main copy, and the "newest" (or some other selection criterion) backup, of identical files, and delete all the spurious extras. This way if a file changes from backup tree to backup tree, even with the same name, I'll have a "change" history. On the other hand it'll clear out all the duplicate copies so that I don't have to search through every one if I'm looking for an older version. Not to mention reducing the amount of backup copies I have lying around.
scp -pr directorytree desktopmachine:backuppartition
I realized that this is taking up a bunch of extra space because it just decends the symlinks like they were in the directory tree.
Is there something easy, simple and convenient for backing up from one machine on a LAN to another? If my house burns down and both machines are cooked, I'm hosed, but short of offsite backups, this seems to be the most expedient way of backing up, if I set it up as a chron job. I can save my desktop data on my server, and my server on my desktop.
I've got a tremndous number of duplicate files, due to brute force backups. At the price of harddrive space, it's not worth the time to go through and cull duplicate files. I'd love something that would keep the main copy, and the "newest" (or some other selection criterion) backup, of identical files, and delete all the spurious extras. This way if a file changes from backup tree to backup tree, even with the same name, I'll have a "change" history. On the other hand it'll clear out all the duplicate copies so that I don't have to search through every one if I'm looking for an older version. Not to mention reducing the amount of backup copies I have lying around.
