Amanda
Gurus,
Fairly new to
amanda, but getting great results so far :)
I'm a little
perplexed by the following output from the balance command of
amadmin:
due-date
#fs orig KB out KB
balance
-------------------------------------------
10/17 Wed 1 7917267 5560000 +270.6%
10/18 Thu 0 0 0 ---
10/19 Fri 0 0 0 ---
10/20 Sat 0 0 0 ---
10/21 Sun 0 0 0 ---
10/22 Mon 15 3804531 1404544 -6.4%
10/23 Tue 1 1760270 537280 -64.2%
-------------------------------------------
TOTAL 17 13482068 7501824 1500364 (estimated 5 runs per dumpcycle)
-------------------------------------------
10/17 Wed 1 7917267 5560000 +270.6%
10/18 Thu 0 0 0 ---
10/19 Fri 0 0 0 ---
10/20 Sat 0 0 0 ---
10/21 Sun 0 0 0 ---
10/22 Mon 15 3804531 1404544 -6.4%
10/23 Tue 1 1760270 537280 -64.2%
-------------------------------------------
TOTAL 17 13482068 7501824 1500364 (estimated 5 runs per dumpcycle)
The man page for
amadmin says that amanda tries to balance the number of full backups across the
dump cycle and that the balance column shows how far off that night's
backups are from the average. It also says amanda tries to keep the backups
within +/- 5%.
Some
particulars:
Amanda 2.4.2p2-4 on
RH 7.1
dumpcycle
7
runspercycle
5
tapecycle
17
disklist contains
between 5-6 disk entries for 3 different hosts, all Amanda 2.4.2p2-4 on RH
7.1.
The filesystem to be backed up tonight
that causes the +270.6% above is the single largest filesystem on all the
machines (7.6GB of 23GB used).
So, I'm I crazy
or is amanda way off in it's balancing? And, more importantly, other than tape
usage, should I be concerned?
TIA,
Doug
