Edgy versions are in compatible with older version

Hello and thank you for reporting. Your initial description lacks the detail necessary to diagnose the problem. Without some sort of error message it will be hard to reproduce and confirm the bug. Please specify the error/info message you get that is indicative of the version incompatibilities.

I have been debugging this further and discovered the problem is not
what I thought it was. I am now reproducing the problem with both the
Ubuntu/Edgy version of amanda and the Debian/Sarge version of amanda.
However, the problem does occur uniquely to the Ubuntu/Edgy client
machine, so it is something unique to either amanda or the filesystems
on that machine.

The specific error message I get to the log file e-mailed to the server
account is:

estimate of level 1 failed: -1

and so all filesystems result in "RESULTS MISSING".

I do not get any error messages to the amanda logs that indicate more
specifically what the problem is.

If I disable the Ubuntu client, everything works correctly.

                              From:
backup <email address hidden>
                                To:
<email address hidden>
                           Subject:
new AMANDA MAIL REPORT FOR November
11, 2006
                              Date:
Sat, 11 Nov 2006 13:21:41 -0500
(EST)

*** THE DUMPS DID NOT FINISH PROPERLY!

*** A TAPE ERROR OCCURRED: [label new07 or new tape not found in rack].
Some dumps may have been left in the holding disk.
Run amflush to flush them to tape.
The next tape Amanda expects to use is: new07.

FAILURE AND STRANGE DUMP SUMMARY:
  fea sdb1 lev -1 FAILED [no estimate]
  fea sdb2 lev -1 FAILED [no estimate]
  fea sda2 lev -1 FAILED [no estimate]
  fea sda1 RESULTS MISSING
  fea hda1 RESULTS MISSING
  fea hda2 RESULTS MISSING
  fea hdb1 RESULTS MISSING
  fea hdb2 RESULTS MISSING
  fea7 sda1 RESULTS MISSING
  fea7 sda5 RESULTS MISSING
  fea7 sda7 RESULTS MISSING
  fea7 sda9 RESULTS MISSING

STATISTICS:
                          Total Full Daily
                        -------- -------- --------
Estimate Time (hrs:min) 0:01
Run Time (hrs:min) 0:01
Dump Time (hrs:min) 0:00 0:00 0:00
Output Size (meg) 0.1 0.0 0.1
Original Size (meg) 0.1 0.0 0.1
Avg Compressed Size (%) -- -- --
(level:#disks ...)
Filesystems Dumped 3 1 2 (1:1 4:1)
Avg Dump Rate (k/s) 1400.0 185.2 2826.1

Tape Time (hrs:min) 0:00 0:00 0:00
Tape Size (meg) 0.0 0.0 0.0
Tape Used (%) 0.0 0.0 0.0
Filesystems Taped 0 0 0
Avg Tp Write Rate (k/s) -- -- --

NOTES:
  planner: disk fea:hda1, estimate of level 1 failed: -1.
  planner: disk fea:sdb2, estimate of level 1 failed: -1.
  planner: disk fea:sdb1, estimate of level 1 failed: -1.
  planner: disk fea:sda2, estimate of level 1 failed: -1.
  planner: disk fea:sda1, estimate of level 1 failed: -1.
  planner: Incremental of fea7:sda1 bumped to level 2.
  planner: fea sda2 20061111 0 [dumps too big, 10699610 KB, full dump
delayed]
  planner: fea sdb2 20061111 0 [dumps too big, 9480560 KB, full dump
delayed]
  planner: fea sdb1 20061111 0 [dumps too big, 9961600 KB, full dump
delayed]
  planner: Full dump of fea:hda1 promoted from 14 days ahead.
  planner: Full dump of fea:...

I have now confirmed this bug to exist on a second Ubuntu/Edgy system
which is my home machine. The home machine is an amanda server & client
and serves only itself. I have had these errors and only level 0
backups since the upgrade to edgy on this machine.

Home machine errors started on 10-22

Work machine errors started about 10-26 (I upgraded to edgy on the work
machine after the home machine upgrade went OK)

I should have paid closer attention. These amanda problems have existed
since the edgy upgrade.

This may be related to bug #32330. I am not sure, but I will try to
find out.

On Fri, 2006-11-10 at 21:49 +0000, Jeremy Teale wrote:
> Hello and thank you for reporting. Your initial description lacks the
> detail necessary to diagnose the problem. Without some sort of error
> message it will be hard to reproduce and confirm the bug. Please specify
> the error/info message you get that is indicative of the version
> incompatibilities.
>
> ** Changed in: amanda (Ubuntu)
> Status: Unconfirmed => Needs Info
>
--
James D. Freels, Ph.D.
Oak Ridge National Laboratory
<email address hidden>
http://www.comsol.com/stories/hfir/

I have now isolated the problem to my home machine running both the
server and client under ubuntu/edgy and totally unrelated to debian. It
has the same problem, so I design a test backup with only 3 small
filesystems. The test is designed to create a level 1 backup, but each
time a failure in the planner occurs and forces a level 0 backup for all
filesystems. This is not acceptable (to have level 0 backups all the
time for every filesystem will not be possible).

I have found a tar error message deep down in the debug log files that
may be the cause of the problem. However, I do not know how to fix it:
(note the error is not recoverable and exiting ... ):

sendsize: debug 1 pid 8706 ruid 34 euid 34: start at Mon Nov 13 11:16:32
2006
sendsize: version 2.5.0p2
sendsize[8708]: time 0.034: calculating for amname 'hdb2', dirname
'/xp', spindle -1
sendsize[8708]: time 0.035: getting size via gnutar for hdb2 level 0
sendsize[8708]: time 0.037: spawning /usr/lib/amanda/runtar in pipeline
sendsize[8708]: argument list: /bin/tar --create --file /dev/null
--directory /xp --one-file-system --numeric-owner
--listed-incremental /var/lib/amanda/gnutar-lists/fea_homehdb2_0.new
--sparse --ignore-failed-read --totals
--exclude-from /tmp/amanda/sendsize.hdb2.20061113111632.exclude .
sendsize[8710]: time 0.039: calculating for amname 'hdb3', dirname
'/mp3s', spindle -1
sendsize[8710]: time 0.039: getting size via gnutar for hdb3 level 0
sendsize[8710]: time 0.041: spawning /usr/lib/amanda/runtar in pipeline
sendsize[8710]: argument list: /bin/tar --create --file /dev/null
--directory /mp3s --one-file-system --numeric-owner
--listed-incremental /var/lib/amanda/gnutar-lists/fea_homehdb3_0.new
--sparse --ignore-failed-read --totals
--exclude-from /tmp/amanda/sendsize.hdb3.20061113111632.exclude .
sendsize[8712]: time 0.042: calculating for amname 'hdb9', dirname
'/iso', spindle -1
sendsize[8712]: time 0.043: getting size via gnutar for hdb9 level 0
sendsize[8706]: time 0.045: waiting for any estimate child: 3 running
sendsize[8712]: time 0.046: spawning /usr/lib/amanda/runtar in pipeline
sendsize[8712]: argument list: /bin/tar --create --file /dev/null
--directory /iso --one-file-system --numeric-owner
--listed-incremental /var/lib/amanda/gnutar-lists/fea_homehdb9_0.new
--sparse --ignore-failed-read --totals
--exclude-from /tmp/amanda/sendsize.hdb9.20061113111632.exclude .
sendsize[8708]: time 0.110: Total bytes written: 296960 (290KiB,
33MiB/s)
sendsize[8708]: time 0.111: .....
sendsize[8708]: estimate time for hdb2 level 0: 0.074
sendsize[8708]: estimate size for hdb2 level 0: 290 KB
sendsize[8708]: time 0.111: waiting for /bin/tar "hdb2" child
sendsize[8708]: time 0.111: after /bin/tar "hdb2" wait
sendsize[8708]: time 0.112: getting size via gnutar for hdb2 level 1
sendsize[8710]: time 0.117: Total bytes written: 10240 (10KiB, 12MiB/s)
sendsize[8712]: time 0.133: Total bytes written: 553697280 (529MiB,
41GiB/s)
sendsize[8712]: time 0.134: .....
sendsize[8712]: estimate time for hdb9 level 0: 0.088
sendsize[8712]: estimate size for hdb9 level 0: 540720 KB
sendsize[8712]: time 0.134: waiting for /bin/tar "hdb9" child
sendsize[8712]: time 0.134: after /bin/ta...

Having isolated the problem, I downgraded the tar package from
v1.15.91-2 normally installed with Ubuntu/Edgy to the older v1.14-2.2
that is installed with the current Debian/Sarge/Stable. This, indeed,
fixed the problem. So, should I file a tar bug report ?

P.S. I did not try the new v1.16-1 that is available via
Debian/Testing/Unstable.

--
James D. Freels, Ph.D.
Oak Ridge National Laboratory
<email address hidden>
http://www.comsol.com/stories/hfir/

Fixed in Feisty (Closed Ubuntu task and added Edgy task)

amanda (1:2.5.0p2-2) unstable; urgency=low
 .
   * merge updated French translation, closes: #369243
   * merge updated Italian translation, closes: #369588
   * merge updated Czech translaterion, closes: #370299
   * merge updated Dutch translaterion, closes: #374850
   * merge updated Swedish translaterion, closes: #375747
   * rework postinst to use adduser instead of usermod to add user backup to
     groups disk and tape, and add postrm support for removing user backup from
     those groups on purge, closes: #380336
   * patch from upstream 2.5.1b1 via Geert Uytterhoeven to support tar 1.15.91,
     closes: #378558
   * accept patch from Raphael Pinson adding xinetd support
   * Clean debian/po/templates.pot after build