-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Paul,
Thanks for your answer.
I had a look in previous report. Last was of the 26th of April and
previous was of the 25th of April :
HOSTN DISK L ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s
- ---------------------- --------------------------------- -----------
s ws/ECOLE-SQL1 1 N/A 0 -- 0:01 0.0 0:00 0.0
s ws/ECOLE-SQL2 0 287460 287460 -- 1:24 3439.0 0:19 15229.6
s ws/ECOLE-VALID 1 N/A 0 -- 0:01 0.0 0:00 0.0
And this one before :
HOSTN DISK L ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s
- ---------------------- --------------------------------- -----------
s ws/ECOLE-SQL1 0 1892430 1892430 -- 8:55 3536.0 1:39 19069.5
s ws/ECOLE-SQL2 0 FAILED --------------------------------------------
s ws/ECOLE-VALID 0 1323040 1323040 -- 6:23 3455.2 1:09 19176.4
Here are they disklist conf and dumptype conf :
iris01bk //win-bck/ECOLE-VALID nocomp-user-gnutar-2100
iris01bk //win-bck/ECOLE-SQL2 nocomp-user-gnutar-2100
iris01bk //win-bck/ECOLE-SQL1 nocomp-user-gnutar-2100
define dumptype nocomp-user-gnutar {
~ program "GNUTAR"
~ comment "Non-root partitions on slow machines"
~ # compress server fast
~ compress none
~ index
}
define dumptype nocomp-user-gnutar-2100 {
~ nocomp-user-gnutar
~ comment "Root partition avec compression a partir de 21H00"
~ starttime 2100
~ priority high
}
In sendsize log file :
sendsize[1034]: time 12.429: calculating for amname 'ws/ECOLE-SQL2',
dirname 'ws/ECOLE-SQL2', spindle -1
sendsize[1034]: time 12.429: getting size via smbclient for
ws/ECOLE-SQL2 level 0
sendsize[1034]: time 12.429: spawning /usr/bin/smbclient in pipeline
sendsize[1034]: argument list: smbclient \\ws\ECOLE-SQL2 -d 0 -U
backup_doc -E -W domaine -c "archive 0;recurse;du"
sendsize[969]: time 12.434: waiting for any estimate child: 1 running
sendsize[1034]: time 12.703:
sendsize[1034]: time 12.703: 43406 blocks of size 2097152.
18666 blocks available
sendsize[1034]: time 12.703: Total number of bytes: 294338560
sendsize[1034]: time 12.704: .....
sendsize[1034]: estimate time for ws/ECOLE-SQL2 level 0: 0.274
sendsize[1034]: estimate size for ws/ECOLE-SQL2 level 0: 287440 KB
sendsize[1034]: time 12.704: waiting for /usr/bin/smbclient
"ws/ECOLE-SQL2" child
sendsize[1034]: time 12.704: after /usr/bin/smbclient "ws/ECOLE-SQL2"
wait
sendsize[1034]: time 12.704: getting size via smbclient for
//win-bck/ECOLE-SQL2 level 1
sendsize[1034]: time 12.704: spawning /usr/bin/smbclient in pipeline
sendsize[1034]: argument list: smbclient ws\ECOLE-SQL2 -d 0 -U
backup_doc -E -W bagnolet.ftmms -c "archive 1;recurse;du"
sendsize[1034]: time 12.984:
sendsize[1034]: time 12.984: 43406 blocks of size 2097152.
18666 blocks available
sendsize[1034]: time 12.984: Total number of bytes: 0
sendsize[1034]: time 12.985: .....
sendsize[1034]: estimate time for ws/ECOLE-SQL2 level 1: 0.280
sendsize[1034]: estimate size for ws/ECOLE-SQL2 level 1: 0 KB
sendsize[1034]: time 12.985: waiting for /usr/bin/smbclient
"ws/ECOLE-SQL2" child
sendsize[1034]: time 12.985: after /usr/bin/smbclient
"//win-bck/ECOLE-SQL2" wait
sendsize[1034]: time 12.985: done with amname 'ws/ECOLE-SQL2',
dirname 'ws/ECOLE-SQL2', spindle -1
sendsize[969]: time 12.985: child 1034 terminated normally
Why does amanda make backup if there is a zero size to dump ?
Paul Bijnens wrote:
| Christian Molière wrote:
|
|> I have strange information in my report file last night :
|
|
| I removed some space from the output a little, and added the
| headers again to make it more readable:
|
| DUMPER STATS TAPER STATS
| HOSTN DISK L ORIG-KB OUT-KB COMP% MMM:SS KB/s MMM:SS KB/s
| ---------------------- --------------------------------- -----------
| serv ws/ECOLE-SQL1 1 N/A 0 -- 0:01 0.0 0:00 0.0
| serv ws/ECOLE-SQL2 1 N/A 0 -- 0:01 0.0 0:00 0.0
| serv s/ECOLE-VALID 1 N/A 0 -- 0:01 0.0 0:00 0.0
|
| There are no DUMP statistics: these entries were flushed from
| holdingdisk.
| Maybe you have "autoflush" on, it that case it was the normal
| dump run that autoflushed the files it found on the holdingdisk.
|
| For the dump itself, you probably have to look of the previous mail
| reports.
|
| From the numbers above, it seems the size of the files was 0 bytes.
| I guess there was only an amanda header of 32Kbytes, which is now
on tape.
|
| The amanda header contains the real date of the dump: you know then
| which report to look at.
|
|> I tried to see what are there in dump file with this command :
|> dd if=00046.serveur.__windows_ECOLE-SQL1.1 bs=32k skip=1 |tar tvf -
|>
|> here it is the output dd command :
|> 0+0 enregistrements lus.
|> 0+0 enregistrements écrits.
|
|
| Yes, an empty file. Have a look at the amanda header itself too.
|
I had a look :
AMANDA: FILE 20040425 s ws/ECOLE-SQL1 lev 1 comp N program
/usr/bin/smbclient
|
|> then amadmin command says there are 46 files in it.
|
|
| Not in it, it is the 46th backup image on the tape.
|
|
|> Does someone know why I have such kind of mistake ?
|
|
| Never seen it myself. Maybe you have more information from the
| report when the dump actually ran. Somehow samba created 0 bytes
| output, but did not signal this as an error. Or something else?
|
|
|> What does mean N/A in report file ?
|
|
| "Not Applicable". Some people say "Not Available".
|
- --
Cordialement,
Sincerely,
Christian MOLIERE
Tél : 01.43.60.11.60
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.1 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFAjPkJZ30NRCZKZ48RAsYQAKDGIdiCWRgiX+m0UrAD0Mg3YJL1AwCbBRii
sZDulnoylf8YRVGnGuSKTWE=
=ls+y
-----END PGP SIGNATURE-----
|
