Post the server/montly/amindexd.*.debug file
Johan Booysen wrote:
On Friday, I tried a full monthly backup using the config as mentioned
in the email below.
The backup seems to have run successfully, and I received the amanda
email showing that all disklist entries had a level 0 backup done, but
amrecover doesn’t seem able to restore from tape, e.g.
amrecover> setdisk /matrix/data1
200 Disk set to /matrix/data1.
*No index records for disk for specified date*
If date correct, notify system administrator
*-sh-3.2$ less client/monthly/amrecover.20110328144524.debug*
1301319924.401732: amrecover: pid 5380 ruid 0 euid 0: start at Mon Mar
28 14:45:24 2011
1301319924.402305: amrecover: pid 5380 ruid 0 euid 0: rename at Mon
Mar 28 14:45:24 2011
1301319924.402376: amrecover: security_getdriver(name=bsdtcp) returns
0x1531e0
1301319924.402417: amrecover: security_handleinit(handle=0x838a058,
driver=0x1531e0 (BSDTCP))
1301319924.402755: amrecover: security_streaminit(stream=0x838d878,
driver=0x1531e0 (BSDTCP))
1301319924.403683: amrecover: connect_port: Skip port 700: owned by epp.
1301319924.404249: amrecover: connect_port: Skip port 701: owned by lmp.
1301319924.404823: amrecover: connect_port: Skip port 702: owned by
iris-beep.
1301319924.408278: amrecover: connect_port: Try port 703: available -
Success
1301319924.408388: amrecover: connected to 127.0.0.1.10080
1301319924.408403: amrecover: our side is 0.0.0.0.703
1301319924.408417: amrecover: try_socksize: send buffer size is 65536
1301319924.408431: amrecover: try_socksize: receive buffer size is 65536
1301319929.464602: amrecover: security_streaminit(stream=0x83958b0,
driver=0x1531e0 (BSDTCP))
1301319929.464640: amrecover: security_close(handle=0x838a058,
driver=0x1531e0 (BSDTCP))
1301319929.464656: amrecover: security_stream_close(0x838d878)
1301319947.583494: amrecover: user command: 'sethost npt01'
1301319949.698629: amrecover: user command: 'listdisk'
1301319959.814081: amrecover: user command: 'setdisk /matrix/data1'
1301320116.448051: amrecover: user command: 'exit'
1301320116.448440: amrecover: security_stream_close(0x83958b0)
1301320116.448492: amrecover: pid 5380 finish time Mon Mar 28 14:48:36
2011
*-sh-3.2$ amadmin monthly find npt01 /matrix/data1*
date host disk lv tape or
file file part status
2011-03-25 20:05:26 npt01 /matrix/data1 0
monthly-3 328 1/1 OK
*-sh-3.2$ cat /etc/amanda/monthly/log.20110325200526.0 | grep data1*
DISK planner npt01 /matrix/data1
INFO planner Adding new disk npt01:/matrix/data1.
SUCCESS dumper npt01 /matrix/data1 20110325200526 0 [sec 113.520 kb
331365 kps 2919.0 orig-kb 1719510]
SUCCESS chunker npt01 /matrix/data1 20110325200526 0 [sec 113.870 kb
326362 kps 2866.4]
STATS driver estimate npt01 /matrix/data1 20110325200526 0 [sec 839
nkb 1719542 ckb 859808 kps 1024]
PART taper monthly-3 328 npt01 /matrix/data1 20110325200526 1/1 0 [sec
6.126424 kb 326361 kps 53271.143702]
DONE taper npt01 /matrix/data1 20110325200526 1 0 [sec 6.126424 kb
326361 kps 53271.143702]
*-sh-3.2$ ls -l /etc/amanda/monthly/index/npt01/_matrix_data1*
total 124
-rw------- 1 amandabackup disk 119158 Mar 26 08:44 20110325200526_0.gz
It looks like using amrestore will work, although it’s still running
at the moment:
*-sh-3.2$ amrestore tape:/dev/nst0 npt01 /matrix/data1 2011-03-25*
…
…
…
amrestore: 210: skipping split dumpfile: date 20110325200526 host
npt01 disk /matrix/data part 119/151 lev 0 comp .gz program /bin/tar
crypt enc server_encrypt /usr/sbin/amgpgcrypt server_decrypt_option -d
amrestore: 211: skipping split dumpfile: date 20110325200526 host
npt01 disk /matrix/data part 120/151 lev 0 comp .gz program /bin/tar
crypt enc server_encrypt /usr/sbin/amgpgcrypt server_decrypt_option -d
Can anyone please help me understand why this would be?
Thanks.
*From:* [email protected]
[mailto:[email protected]] *On Behalf Of *Johan Booysen
*Sent:* 15 March 2011 12:25
*To:* [email protected]
*Subject:* monthly archive config
I’d really appreciate it if someone can just go over what I’m doing
below, so I don’t miss out anything silly.
I’m using Amanda version 2.6.0p1, and finally cannot effectively make
full backups each night anymore, so I’m looking into using Amanda the
‘classic’ way and to set up another config for monthly full backups.
If I recall correctly, then something like this should work for the
monthly backup runs. I simply made a copy of my daily config and am
just modifying the existing files, e.g. amanda.conf:
dumpcycle 0 to make full backups on each run
runspercycle 1 not really sure about this because
dumpcycle is 0??
tapecycle 12 tapes one tape for each month of the year, but
will actually use 15 in the set up tapes for the odd amflush, if necessary
index yes to generate the catalogue for
use by amrecover
record no apparently: “This is normally
enabled for daily backups and turned off for periodic archival runs.”
(Amanda wiki)
infofile "/etc/amanda/monthly/curinfo"
logdir "/etc/amanda/monthly"
indexdir "/etc/amanda/monthly/index"
Other than that, because I made a copy of the daily config, I’ve
cleaned out the copied config folder’s index and log files, and
tapelist file.
Does that look sensible?
Thanks.
Johan
