On Wed, Sep 10, 2003 at 12:48:35PM +0200, Mark Le Noury wrote:I'm just a bit worried, for example, that the full dump for 1 disk on 1 server gets done on a Monday, while the full dump for another disk on another server gets done on a Tuesday etc.
Why does this worry you? As long as you know that every disk will get at least one full dump per dumpcycle, what difference does it make if they're not all on the same tape?
Indeed. It may look strange at first sight, but even in reality this works out fine very well.
I have five disks per server, 15 in all. I would like to be able to
pick up a tape and say "this is a tape with a full dump on it" or even "this is a tape with a few full dumps on it".
Well, that's easy! With 15 disks, it's pretty much guaranteed that amanda will put at least one full dump on each and every tape, so you
can pick up _any_ of your tapes and say, "This is a tape with a few full dumps on it."
:- :-) :-)
a bit of a management nightmare.
Not at all. First there is as Dave said the indexes. But even without indexes you still have your "amadmin Config info" to find out which tape is which. And if that disk is crashed too, you still have the paper postscript labels that amanda prints for you that fit nicely in the tape box. My printed labels contain:
FileNr, Hostname, FilesystemName, Level, OriginalSize, CompressedSize
for each of the dumps, besides of course the Tapename, the Date, the Amanda version, and some general statistics. And it folds nicely in the tape box.
If you really lost everything then you just look trough the printed labels, find the last level 0 of a host/disk and start restoring, followed by the last level 1 etc. Just tried it, and it takes less than 30 seconds to find the correct tapes, *without* computer help.
If you're unfortunate that even the paper inserts got lost (how unfortunate can you get :-) ) then the tape itself still contains for each file a header with host/filesystem/level and date and even the command you need to restore it.
On a general note, you need backups for two different things: user errors and disasters.
One thing is user errors etc, the backups to restore one or a few files that were messed up / deleted by accident. Because you want to respond to these kind of request fast enough, it's not a good idea to put these tapes somewhere in a safe offsite. I run a "daily" config with amanda for this and can go back up to a month and the tapes are stored in the computer room.
The other thing is disasters, and for this you have to calculate what data do you allow to lose given cost/time constraints. I run a weekly "archive" configuration with full backups only in the weekends, and those tapes are stored in a safe, even a few kilometers away. (yes, with printed labels in each tape box.) Some of these archive tapes are marked "no-reuse" now and then and the other one are recycled after a sufficient time. By marking some tapes no-reuse, they keep out the normal cycle that amanda expects, and by reusing the other ones, I don't make this company bankrupt by my supply of tapes. I can go back finding files up to 1999 with this config (I started the archive config in 1999 - the previous archives are not run with Amanda, and they are a lot more difficult to find out what is what).
Not much of a nightmare, if you ask me. On the contrary, I sleep better knowing that she's got my back.
:-)
-- Paul Bijnens, Xplanation Tel +32 16 397.511 Technologielaan 21 bus 2, B-3001 Leuven, BELGIUM Fax +32 16 397.512 http://www.xplanation.com/ email: [EMAIL PROTECTED] *********************************************************************** * I think I've got the hang of it now: exit, ^D, ^C, ^\, ^Z, ^Q, F6, * * quit, ZZ, :q, :q!, M-Z, ^X^C, logoff, logout, close, bye, /bye, * * stop, end, F3, ~., ^]c, +++ ATH, disconnect, halt, abort, hangup, * * PF4, F20, ^X^X, :D::D, KJOB, F14-f-e, F8-e, kill -1 $$, shutdown, * * kill -9 1, Alt-F4, Ctrl-Alt-Del, AltGr-NumLock, Stop-A, ... * * ... "Are you sure?" ... YES ... Phew ... I'm out * ***********************************************************************
