Urs Forster wrote:
Ok, I deletet all indexes, all curinfo, removed all extra tapes from
changer.conf,
ran amrmtape, amtape update, amcheckdb and I found out what fell swoop means
(rather than swell foop; that was the harder part ;-). Despite all of those,
amoverview still shows around 90 days across.
Any more ideas?
Did you also clean out the holdingdisk directory?
I believe a command like:
amadmin YourConfig find --sort d | head
will quickly show which entry is around 90 days old. Probably something
left in your holdingdisk.
I'm doing backup to disk (tapeless). In order to fit the largest filesystem
I had to define the 'tape'-size to be around 3 Gb. Now, amanda seems to be
a bit greedy - meaning it tries to fill every tape (by promoting full dumps).
Since all 'tapes' lie on the same disk, I ran into space problems. Therefore
I had to go back to 5! (until I can overcome the diskspace-shortage).
Could one prevent amanda from promoting fulldumps? (like only one fulldump
in a cycle, even thoug the tape is nearly emtpy?
For promoting full dumps, amanda does not look at the tapesize (except
to see if it will fit). Amanda tries to level out the amount of full
dumps over the runs in a tapecycles.
The command "amadmin YourConfig balance" will tell you how much big
the full dumps are in total, and how much amanda will do each run.
From one of my configs:
$ amadmin daily balance
due-date #fs orig KB out KB balance
----------------------------------------------
3/29 Mon 35 52076537 21884417 -13.0%
3/30 Tue 47 65993395 25774163 +2.5%
3/31 Wed 26 45819482 25694830 +2.2%
4/01 Thu 17 53597714 27243524 +8.3%
4/02 Fri 0 0 0 ---
4/03 Sat 0 0 0 ---
4/04 Sun 0 0 0 ---
----------------------------------------------
TOTAL 126 217487722 100597534 25149383
DISTINCT 123 217485940 100595734
There is about 100 Gbyte of data (after compression) to be spread
over 4 runs in a week, giving about 25 Gbyte each run.
(The fact that on monday, there was 13% less, was a result
of a failed computer.)
Amanda tries to level out the full dumps so that each
run is about the same size, i.e. 25149383 Kbytes. What is
not run as full dump is done with an incremental level, which
has his own method (bumpsize, bumpdays, etc) to tune.
If the tape is too small, amanda will delay full dumps, but if
the tape has plenty of room left, amanda will not promote dumps
just because. If the balance is good enough, there will be
no promotions.
Conclusion: there must be some other reason than changing the
tape-length. If the tapes are filled around 100% each day,
then you probably configured amanda with a tapecycle too low
for that tape-capacity. In that case, amanda tries desperately
to find her balance, by promoting and delaying the full dumps,
but she will never succeed completely.
Bitte keine Antwort an diese Mailadresse. Wird nicht gelesen.
Don't reply to this address.
--
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: Paul.Bijnens AT xplanation DOT com
***********************************************************************
* 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 *
***********************************************************************
|