amstatus clearly tell you it run one dumper because of 'no-bandwidth'.
You must increase 'netusage' and interface 'use' parameter.

'amadmin daily info' will tell the bandwidth (dump rates) for each dle.

Jean-Louis

Charles Stroom wrote:

Hi all,

Amanda is running fine on my home pc ("fiume") with 1 other pc
connected as well ("stremen").  However there are 3 questions which I
can maybe illustrate with the results of amstatus while the backup was
running.

fiume:# su - amandabackup -c "exec amstatus daily"
Using /var/log/amanda/daily/amdump from Thu Nov 15 11:05:01 CET 2007

fiume.localnet:/ 0 74343k wait for dumping fiume.localnet:/home 1 181008k wait for dumping fiume.localnet:/home/charles/pictures 1 1830k finished (11:49:03) fiume.localnet:/home/charles/vmware-4 1 395782k wait for dumping fiume.localnet:/local 0 32k wait for dumping fiume.localnet:/opt 0 487120k wait for dumping fiume.localnet:/usr 0 1651873k finished (11:40:49) fiume.localnet:/usr/local 0 284697k wait for dumping fiume.localnet:/var 0 352587k wait for dumping stremen.localnet:/ 0 1096018k dumping 1044416k

( 95.29%) (11:12:37) stremen.localnet:/srv                 0       480k
finished (11:49:05) stremen.localnet:/var                 0    469264k
finished (11:48:49)

SUMMARY          part      real  estimated
                           size       size
partition       :  12
estimated       :  12              5003163k
flush           :   0         0k
failed          :   0                    0k           (  0.00%)
wait for dumping:   7              1775569k           ( 35.49%)
dumping to tape :   0                    0k           (  0.00%)
dumping         :   1   1044416k   1096018k ( 95.29%) ( 20.88%)
dumped          :   4   2123447k   2131576k ( 99.62%) ( 42.44%)
wait for writing:   0         0k         0k (  0.00%) (  0.00%)
wait to flush   :   0         0k         0k (100.00%) (  0.00%)
writing to tape :   0         0k         0k (  0.00%) (  0.00%)
failed to tape  :   0         0k         0k (  0.00%) (  0.00%)
taped           :   4   2123447k   2131576k ( 99.62%) ( 42.44%)
  tape 1        :   4   2123447k   2131576k ( 17.76%) daily-08
1 dumper idle   : no-bandwidth
taper writing, tapeq: 0
network free kps:      2696
holding space   : 106966008k ( 98.99%)
chunker0 busy   :  0:05:20  ( 12.11%)
chunker1 busy   :  0:17:02  ( 38.67%)
 dumper0 busy   :  0:41:47  ( 94.85%)
 dumper1 busy   :  0:17:02  ( 38.67%)
   taper busy   :  0:36:26  ( 82.71%)
 0 dumpers busy :  0:07:45  ( 17.59%)            not-idle:  0:07:35
( 97.99%) no-bandwidth:  0:00:09  (  2.01%)
 1 dumper busy  :  0:36:18  ( 82.40%)        no-bandwidth:  0:36:18
(100.00%) 2 dumpers busy :  0:00:00  (  0.00%)
fiume:/usr/local/flexlm/licenses # date
Thu Nov 15 11:52:38 CET 2007

The first question is why the dumping on "stremen" is soooo slow!  It
did 95% of ~1Gb in 40 minutes.  Dumping on fiume (the one which runs
the server) is much faster, maybe 10x faster.  Even downloading from
internet goes faster than the dumps on stremen.

The second question is that always only 1 dumper seems to be running at
the time.  Even if the dumper runs on stremen, fiume will not start
another dumper in parallel.  I thought that I specified sufficient
bandwidth and # of dumpers in the conf file, which is attached.

And why are small DLEs written to tape, when I have specified
dumporder "Ss" in the conf file?

Just questions, nothing really important, but they do intrigue me.

Thanks for any clue.

Charles