How to manage your backups

Access to the backup server, RSBKSRV, is via ssh using public/private keys. This is general shell access, and you have an ordinary home directory on that machine.

If you need to create such a key pair, simply run

    ssh-keygen

Please send a ssh public key (not a private key -- the filename should have ".pub" in its name) to 'langley@cs.fsu.edu'. I will install this on the server and this will let you access your area on the backup server.

You can use whatever method you like to put data into your home directory on RSBKSRV. I recommend that you use rsync with separate repository directories for each input source that you have. (Please note that rsync now uses ssh by default; the link to rsync doesn't actually make that explicit, but it has been the default behavior rsync program for several years now.) Note that once you are able to login, you can modify ~/.ssh/authorized_keys to include whatever other public keys you like.

Backup examples

For instance, let's suppose that you have three machines that you have data that you want to back up: server1, server2, and server3. On server1, you have three subdirectories that you want backed up, and on the other two machines you have one each:

    server1: /home/me, /good/stuff, /real/good/stuff
    server2: /home/myself
    server3: /home/superlative

I would recommend creating the following directories on rsbksrv, assuming that your home directory on rsbksrv is /home/faculty/me:

    /home/faculty/me/server1/home/me/          # server1's /home/me repository
    /home/faculty/me/server1/good/stuff/       # server1's /good/stuff repository
    /home/faculty/me/server1/real/good/stuff/  # server1's /real/good/stuff repository

    /home/faculty/me/server2/home/myself/      # server2's /home/myself repository

    /home/faculty/me/server3/home/superlative/ # server3's /home/superlative repository

Then you could back these up on each machine as simply as

    me@server1% rsync -a /home/me/ /good/stuff/ /real/good/stuff/ rsbksrv:server1/

    me@server2% rsync -a /home/myself/ rsbksrv:server2/

    me@server3% rsync -a /home/superlative/ rsbksrv:server3/

If you want to be a more discriminating (for instance, you probably have a number of subdirectories that you might want to exclude), you can do quite a number of things; for instance, you probably don't want to also send your ssh keys, so you might do something like:

    rsync -v -a --exclude='.ssh/id_rsa*' --exclude=.ssh/.old-id_rsa /home/me/ rsbksrv:server1/

Please note that it rsync works well coming from a Windows environment, so you can also arrange backups from a Windows client or server machine.

The tape window

You might want to consider your timing of backups: we start write tapes each day starting at 3:00am; it would probably be best to leave your files quiescent during the time that taping is going on.

Other considerations

If you are backing up a database that uses a binary format, it might be a good idea to dump it to text format before you commence your backup. The reason is that the database may well be acting on that binary image during the time that you are backing it up.

This is also true for other programs that store data in a binary format that is operated directly on by a persistent process.

Presently, each person's quota space is limited to 28 gigabytes.