did the amdump run successfully?

1) go to backup-dir/data
2) look for the file 00001.myhost.mynetwork.com._etc_amanda.0
3) what does the header of the above file say?
  it should be something similar to

To restore, position tape at start of file and run:
        dd if=<tape> bs=32k skip=1 | /usr/local/sbin/amgtar -f ... -
^L



Ong Loeng Seng wrote:

Hi Kevin,

I did run multiple daily dumps in a day but it was okay before I changed
amgtar permission etc.

I'll run once a day and see if there is any changes after 5 days of full
dumps.

Following is the latest amtrmidx:
/tmp/amanda/amtrmidx.20060109170452.debug

amtrmidx: debug 1 pid 8319 ruid 33 euid 33: start at Mon Jan  9 17:04:52
2006
/usr/local/libexec/amtrmidx: version 2.4.5p1
myhost.mynetwork.com /etc/amanda
myhost.mynetwork.com /home/SCOUTNET/loeng.ong/Desktop
amtrmidx: pid 8319 finish time Mon Jan  9 17:04:53 2006
You have new mail in /var/spool/mail/root

/var/log/amanda/amindexd file: amindexd.20060108094717.debug

amindexd: debug 1 pid 3825 ruid 33 euid 33: start at Sun Jan  8 09:47:17 2006
amindexd: version 2.4.5
amindexd: time 0.000: < 220 myhost AMANDA index server (2.4.5) ready.
amindexd: time 0.001: > SECURITY USER root
amindexd: time 0.001: bsd security: remote host myhost.mynetwork.com user root 
local user amanda
amindexd: time 0.001: amandahosts security check passed
amindexd: time 0.001: < 200 Access OK
amindexd: time 0.041: > FEATURES fffffeff9ffe7f
amindexd: time 0.041: < 200 FEATURES fffffeff9ffe7f
amindexd: time 0.081: > DATE 2006-01-08
amindexd: time 0.081: < 200 Working date set to 2006-01-08.
amindexd: time 0.121: > SCNF DailySet1
amindexd: time 0.123: < 200 Config set to DailySet1.
amindexd: time 0.162: > HOST myhost.mynetwork.com
amindexd: time 0.162: < 200 Dump host set to myhost.mynetwork.com.
amindexd: time 0.203: > DISK /
amindexd: time 0.203: < 501 Disk myhost.mynetwork.com:/ is not in your disklist.
amindexd: time 0.243: > DISK rootfs
amindexd: time 0.243: < 501 Disk myhost.mynetwork.com:rootfs is not in your 
disklist.
amindexd: time 57.763: > HOST myhost.mynetwork.com
amindexd: time 57.763: < 200 Dump host set to myhost.mynetwork.com.
amindexd: time 65.261: > DISK /etc/amanda
amindexd: time 65.261: < 200 Disk set to /etc/amanda.
amindexd: time 65.301: > OISD /
amindexd: time 65.301: < 500 No dumps available on or before date "2006-01-08"
amindexd: time 1814.577: > QUIT
amindexd: time 1814.578: < 200 Good bye.
amindexd: time 1814.578: pid 3825 finish time Sun Jan  8 10:17:31 2006

I am confused now. my disklist is recognised in amtrmidx file but not in
amindexd file ??

-- Loeng


On Mon, 2006-01-09 at 09:57 -0800, Kevin Till wrote:

Ong Loeng Seng wrote:

Hi Kevin,

Hi Ong Loeng,

I have done quite a bit of amrecover testing on 2.4.5 and I know it works very well.

In /tmp/amanda directory, find the latest amindexd*.debug see if there is any error after the line

amindexd: uncompress command: /usr/bin/gzip -dc ......



--Kevin

It does say that it can contact the index server.
This is the output when I run "amrecover -C DailySet1"

AMRECOVER Version 2.4.5p1. Contacting server on myhost.mynetwork.com ...
220 myhost AMANDA index server (2.4.5) ready.
200 Access OK
Setting restore date to today (2006-01-08)
200 Working date set to 2006-01-08.
Warning: no log files found for tape DailySet104 written 2006-01-02
Warning: no log files found for tape DailySet103 written 2006-01-02
Warning: no log files found for tape DailySet102 written 2006-01-02
Warning: no log files found for tape DailySet101 written 2006-01-02
Warning: no log files found for tape DailySet105 written 2006-01-02
Scanning /var/tmp...
200 Config set to DailySet1.
200 Dump host set to myhost.mynetwork.com.
Trying disk / ...
Trying disk rootfs ...
Can't determine disk and mount point from $CWD '/root'
amrecover> sethost myhost.mynetwork.com
200 Dump host set to myhost.mynetwork.com.
amrecover> setdisk /etc/amanda
200 Disk set to /etc/amanda.
No index records for disk for specified date
If date correct, notify system administrator


What do you think?

On Fri, 2006-01-06 at 15:34 -0800, Kevin Till wrote:

did amrecover connect to the right index server?

Do you see something similar to the following?

AMRECOVER Version 2.4.5p1. Contacting server on ... ...
220 boston AMANDA index server (2.4.5p1) ready.
200 Access OK



Ong Loeng Seng wrote:

Hi Paul,

I understand now. Thanks.

However, I 've got another problem : when I run amrecover -C "DailySet1"

...
Setting restore date to today (2006-01-05)
200 Working date set to 2006-01-05.
Scanning /var/tmp...
200 Config set to DailySet1.
200 Dump host set to myhost.mynetwork.com.
Trying disk / ...
Trying disk rootfs ...
Can't determine disk and mount point from $CWD
'/usr/local/amanda-2.4.5p1'
amrecover> setdisk /etc/amanda
200 Disk set to /etc/amanda.
No index records for disk for specified date
If date correct, notify system administrator

My amanda configuration, includes this ...
...
indexdir "/var/lib/amanda/DailySet1/index"      # index directory

define dumptype hard-disk-tar {
  comment "Back up to hard disk instead of tape - using tar"
  holdingdisk no
  index yes
  priority high
  program "GNUTAR"
  strategy noinc
}

and my disklist file content:
myhost.mynetwork.com /etc/amanda hard-disk-tar
...

This is the content
of /var/lib/amanda/DailySet1/index/myhost.mynetwok.com/_etc_amanda
-rw-------  1 amanda disk  115 Jan  5 16:50 20060105_0.gz

It was okay before I do chmod 755 to amaespipe and amgtar files Following are the files:

-rwxr-xr-x   1 root root   2244 Jan  5 12:06 amaespipe
-rwxr-xr-x   1 root root    554 Jan  5 12:04 amgtar

I thought it was a script, so I need to do chmod, am I right? Because
amanda was complaining about GNUTAR program, which should execute amgtar
instead of tar. Since then, it got index problem. I don't understand
why?
Then I thought I might need to reconfigure and recompile amanda, which I
did, but the problem still persists.

Did I miss something?


On Tue, 2006-01-03 at 12:53 +0100, Paul Bijnens wrote:

Ong Loeng Seng wrote:

amrecover> settape
myhost.mynetwork.com:amanda.conf:/var2/amandadumps/tape02
Using tape "amanda.conf:/var2/amandadumps/tape02" from server
myhost.mynetwork.com.

What is this???   Is that the name of your tape device???

When you are using the "FILE" driver, the name is:
myhost.mynetwork.com:file:/var2/amandadumps/tape02

(where "tape02" is the parent of the "data" subdirectory)


[...]

Continue [?/Y/n/s/t]? Y
EOF, check amidxtaped.<timestamp>.debug file on myhost.mynetwork.com
amrecover: short block 0 bytes
UNKNOWN file
amrecover: Can't read file header

Indeed, that's exactly what is to be expected.

[...]

amidxtaped: time 0.000: > DEVICE=disklist:/var2/amandadumps/tape02

THis time, you tried something else!  This file is not from the same
run as the script above.  But it is wrong just as well.

The backup file is definitely in /var2/amandadumps/tape02 directory. I don't understand why amrestore says no such file or directory?

If /var2/amandadumps/tape02 has a subdirectory named "data" which
contains the backup file, then the syntax is:

 settape  myhost.mynetwork.com:file:/var2/amandadumps/tape02

The word "file" in the middle is not a placeholder, but is needed
litterally.


If you are using the chg-disk changer, then it is best to add these
lines to amanda.conf:

amrecover_changer "changer"
amrecover_do_fsf true
amrecover_check_label true

and then you can just do from within amrecover:

settape changer


See:  http://wiki.zmanda.com/index.php/File_driver

--
Thank you!
Kevin Till

Amanda documentation: http://wiki.zmanda.com
Amanda forums:        http://forums.zmanda.com