I'm trying to figure out why may backups and estimages take so long. I have a dual-processor machine with 4GB RAM. / and /boot are two software RAID1 drives and /home is a RAID5 hardware drive. My tape drive is an Overland Neo2000 LTO2 with 30 tape library.
I am using a 200GB holding disk on an isolated drive dedicated to amanda.
The estimate phase on localhost (my backup server) took two hours. I watched the tar process and it took anywhere from 5-30% of the cpu
here is a sample of my amanda report: STATISTICS: Total Full Daily -------- -------- -------- Estimate Time (hrs:min) 2:01 Run Time (hrs:min) 2:47 Dump Time (hrs:min) 1:10 0:17 0:53 Output Size (meg) 6532.4 4628.1 1904.2 Original Size (meg) 6532.4 4628.1 1904.2 Avg Compressed Size (%) -- -- -- (level:#disks ...) Filesystems Dumped 10 5 5 (1:4 3:1) Avg Dump Rate (k/s) 1597.4 4656.7 615.1
Tape Time (hrs:min) 0:04 0:03 0:01 Tape Size (meg) 6532.6 4628.2 1904.3 Tape Used (%) 3.2 2.2 0.9 (level:#disks ...) Filesystems Taped 10 5 5 (1:4 3:1) Avg Tp Write Rate (k/s) 25126.8 25105.0 25179.7
USAGE BY TAPE: Label Time Size % Nb gauss-21 0:04 6532.6 3.2 10
FAILED AND STRANGE DUMP DETAILS:
/-- localhost / lev 3 STRANGE sendbackup: start [localhost:/ level 3] sendbackup: info BACKUP=/bin/tar sendbackup: info RECOVER_CMD=/bin/tar -f... - sendbackup: info end ? gtar: ./tmp/screens/S-root/16229.pts-6.master: Warning: Cannot stat: No such file or directory | gtar: ./tmp/ssh-XXHe66oR/agent.26555: socket ignored | gtar: ./tmp/ssh-XXXjwuZQ/agent.16145: socket ignored | gtar: ./tmp/ssh-XXtUIuYn/agent.12543: socket ignored ? gtar: ./var/lib/amanda/DailySet1/index/mathweb/_home/20030617_0.gz.tmp: Warning: Cannot stat: No such file or directory ? gtar: ./var/lib/ganglia/rrds/Math Gauss Cluster/node5.cluster/load_one.rrd: file changed as we read it | Total bytes written: 51947520 (49MB, 82kB/s) sendbackup: size 50730 sendbackup: end \--------
NOTES: planner: Incremental of localhost:/ bumped to level 3. planner: Full dump of mathweb:/home promoted from 3 days ahead. planner: Full dump of mathweb:/usr/local promoted from 3 days ahead. planner: Full dump of mathweb:/www promoted from 3 days ahead. planner: Full dump of localhost:/boot promoted from 3 days ahead. planner: Full dump of mathweb:/var/log promoted from 3 days ahead. taper: tape gauss-21 kb 6689664 fm 10 [OK]
DUMP SUMMARY:
DUMPER STATS
TAPER STATS
HOSTNAME DISK L ORIG-KB OUT-KB COMP% MMM:SS
KB/s MMM:SS KB/s
----------------------------------
--------------------------------------------- -----------------
localhost / 3 50730 50752 -- 10:18
82.1 0:02 21829.3
localhost /boot 0 13010 13024 -- 0:01
10048.7 0:03 3919.4
localhost /home 1 205640 205664 -- 35:26
96.7 0:10 20045.5
mathweb / 1 68530 68544 -- 1:25
810.3 0:05 12602.2
mathweb /boot 1 10 32 -- 0:00
79.9 0:02 14.8
mathweb /home 0 3138930 3138944 -- 6:29
8071.3 1:50 28619.6
mathweb /usr/local 0 1135110 1135136 -- 9:28
1997.6 0:49 23216.5
mathweb /var/lib/pgsql 1 1625000 1625024 -- 5:41
4770.4 0:57 28380.8
mathweb /var/log 0 212930 212960 -- 0:23
9281.4 0:10 20687.6
mathweb /www 0 239240 239264 -- 0:36
6579.7 0:17 14419.5(brought to you by Amanda version 2.4.4)
Thanks, Jason Edgecombe
