Public bug reported:

When there is more backup output data to store than fits on the (virtual) tape, 
Amanda hangs.
The expected behavior is to terminate the backup, and email (when launched from 
crond) the backup report.

Relevant part of the "ps aux" output:

    backup   2749906  0.0  0.0   2800  1664 ?        S    00:45   0:00 sh -c cd 
/tmp && /usr/sbin/amdump DailySet1
    backup   2749907  0.0  0.0  53620 31328 ?        S    00:45   0:00 
/usr/bin/perl /usr/sbin/amdump DailySet1
    backup   2749909  0.0  0.0  31692 17792 ?        S    00:45   0:00 
/usr/lib/amanda/driver DailySet1 --log-filename 
/var/log/amanda/DailySet1/log.20250223004501.0
    backup   2749910  0.2  0.1 220000 53744 ?        S    00:45   1:10 
/usr/bin/perl /usr/lib/amanda/taper DailySet1 --storage DailySet1 
--log-filename /var/log/amanda/DailySet1/log.20250223004501.0
    backup   2749911  0.2  0.0      0     0 ?        Z    00:45   1:09 [dumper] 
<defunct>
    backup   2749912  0.0  0.0  26136 12928 ?        S    00:45   0:00 dumper1 
DailySet1 --log-filename /var/log/amanda/DailySet1/log.20250223004501.0
    backup   2749913  0.0  0.0  26136 12800 ?        S    00:45   0:00 dumper2 
DailySet1 --log-filename /var/log/amanda/DailySet1/log.20250223004501.0
    backup   2749914  0.0  0.0  26136 12928 ?        S    00:45   0:00 dumper3 
DailySet1 --log-filename /var/log/amanda/DailySet1/log.20250223004501.0
    backup   2752623  0.0  0.0  99548 13056 ?        Ssl  01:21   0:18 amandad 
-auth=bsdtcp amdump amindexd amidxtaped
    backup   2752624  0.0  0.0 173036 12800 ?        Sl   01:21   0:31 
/usr/lib/amanda/sendbackup amandad bsdtcp --shm-name 
/amanda_shm_control-2752623-0
    backup   2752625  0.0  0.0      0     0 ?        Z    01:21   0:00 
[amandad] <defunct>
    backup   2752626  3.3  0.0   3424  1536 ?        S    01:21  17:48 
/usr/bin/gzip --fast
    backup   2752628  0.1  0.0  25040  3304 ?        S    01:21   0:56 
/usr/lib/amanda/sendbackup amandad bsdtcp --shm-name 
/amanda_shm_control-2752623-0
    backup   2752632  0.0  0.0   2800  1792 ?        S    01:21   0:00 sh -c -- 
/usr/bin/tar -tf - 2>/dev/null | sed -e 's/^\.//'
    backup   2752634  0.0  0.0   5804  3328 ?        S    01:21   0:20 
/usr/bin/tar -tf -
    backup   2752635  0.0  0.0   3604  1920 ?        S    01:21   0:00 sed -e 
s/^\.//
    backup   2752636  0.0  0.0      0     0 ?        Z    01:21   0:00 [sh] 
<defunct>

Output of "pstree -pl backup":

    sh(2749906)───amdump(2749907)───driver(2749909)─┬─dumper(2749911)
                                                    ├─dumper(2749912)
                                                    ├─dumper(2749913)
                                                    ├─dumper(2749914)
                                                    └─taper(2749910)

    amandad(2752623)─┬─amandad(2752625)
                     ├─sendbackup(2752624)─┬─gzip(2752626)
                     │                     
├─sendbackup(2752628)───sh(2752632)─┬─sed(2752635)
                     │                     │                                   
└─tar(2752634)
                     │                     ├─tar(2752629)───sh(2752636)
                     │                     ├─{sendbackup}(2752630)
                     │                     └─{sendbackup}(2752631)
                     └─{amandad}(2752627)


Tail of /var/log/amanda/DailySet1/log.20250223004501.0 (host and dle names 
redacted):

    INFO taper tape DAILY06632 kb 81915952 fm 122 [OK]
    ERROR taper no-tape config [1 tapes filled; runtapes=1 does not allow 
additional tapes]
    PARTIAL taper "ST:DailySet1" "POOL:DailySet1" <host> <dle> 20250223004501 1 
0 00000000:0 00000000:0 00000000:0 [sec 1330.000000 bytes 40129789952 kps 
29465.600000] "1 tapes filled; runtapes=1 does not allow additional tapes"
    WARNING driver dumper0 pid 2749911 is messed up, ignoring it.
    WARNING driver dumper0 died while dumping <host>:<dle> lev 0.

After manually killing a few processes 
(2752636/2752635/2752634/2752632/2752628/2749914/2749913/2749912/2749911/2749910),
 Amanda continues.
The following is logged to /var/log/amanda/DailySet1/log.20250223004501.0:

    WARNING driver Taper protocol error
    WARNING driver going into degraded mode because of taper component error.
    FAIL driver <host> <dle> 20250223004501 0 [can't do degraded dump without 
holding disk]
    WARNING driver taper0 pid 2749910 exited with signal 15
    WARNING driver dumper0 pid 2749911 exited with signal 11
    WARNING driver dumper1 pid 2749912 exited with signal 15
    WARNING driver dumper2 pid 2749913 exited with signal 15
    WARNING driver dumper3 pid 2749914 exited with signal 15
    FINISH driver date 20250223004501 time 34579.843
    INFO driver pid-done 2749909
    INFO amdump pid-done 2749907

and Amanda emails the backup report, as before.
The above has been happening for a few days, requiring manual intervention 
after each backup run.
If all backup output data fits on the provided tape, everything works fine.

For comparison, below is the tail of
/var/log/amanda/DailySet1/log.20241218004501.0, the last successful run
of Amanda where data didn't fit on the tape (host and dle names
redacted):

    INFO taper tape DAILY06572 kb 81916048 fm 119 [OK]
    ERROR taper no-tape config [1 tapes filled; runtapes=1 does not allow 
additional tapes]
    INFO dumper pid-done 383967
    PARTIAL taper "ST:DailySet1" "POOL:DailySet1" <host1> <dle1> 20241218004501 
1 0 00000000:0 00000000:0 00000000:0 [sec 273.000000 bytes 20922073088 kps 
74841.435897] "1 tapes filled; runtapes=1 does not allow additional tapes"
    INFO dumper pid-done 383967
    FAIL dumper <host1> <dle1> 20241218004501 0 [data read: recv error: 
shm_ring is cancelled]
      sendbackup: start [<host1>:<dle1> level 0]
      sendbackup: info BACKUP=/usr/bin/tar
      sendbackup: info RECOVER_CMD=/usr/bin/tar -xpGf - ...
      sendbackup: info end
    FAIL driver <host1> <dle2> 20241218004501 0 [can't do degraded dump without 
holding disk]
    FAIL driver <host1> <dle3> 20241218004501 0 [can't do degraded dump without 
holding disk]
    FAIL driver <host2> <dle4> 20241218004501 0 [can't do degraded dump without 
holding disk]
    FAIL driver <host1> <dle1> 20241218004501 0 [can't do degraded dump without 
holding disk]
    INFO dumper pid-done 381454
    INFO dumper pid-done 381456
    INFO dumper pid-done 381457
    INFO dumper pid-done 381455
    INFO taper pid-done 381453
    FINISH driver date 20241218004501 time 2024.541
    INFO driver pid-done 381452
    INFO amdump pid-done 381450

Last update of the Amanda packages was during the upgrade from 20.04LTS
to 24.04LTS:

    2024-12-29 19:48:20 upgrade amanda-server:amd64 1:3.5.1-8ubuntu1.4 
1:3.5.1-11.1build6
    2024-12-29 19:49:36 upgrade amanda-common:amd64 1:3.5.1-8ubuntu1.4 
1:3.5.1-11.1build6
    2024-12-29 19:49:37 upgrade amanda-client:amd64 1:3.5.1-8ubuntu1.4 
1:3.5.1-11.1build6

i.e. after the last sucessful run, and before the first seen failure.

Thanks!
(Amanda user since 1997)

ProblemType: Bug
DistroRelease: Ubuntu 24.04
Package: amanda-server 1:3.5.1-11.1build6
ProcVersionSignature: Ubuntu 6.8.0-51.52-generic 6.8.12
Uname: Linux 6.8.0-51-generic x86_64
ApportVersion: 2.28.1-0ubuntu3.3
Architecture: amd64
CasperMD5CheckResult: unknown
Date: Sun Feb 23 11:51:14 2025
InstallationDate: Installed on 2014-05-18 (3934 days ago)
InstallationMedia: Ubuntu 14.04 LTS "Trusty Tahr" - Release amd64 (20140417)
RebootRequiredPkgs: Error: path contained symlinks.
SourcePackage: amanda
UpgradeStatus: Upgraded to noble on 2024-12-29 (56 days ago)
modified.conffile..etc.default.apport: [modified]
mtime.conffile..etc.default.apport: 2014-07-26T14:08:18.443558

** Affects: amanda (Ubuntu)
     Importance: Undecided
         Status: New


** Tags: amd64 apport-bug noble

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2099800

Title:
  Amanda hangs when tape is filled and no additional tapes are available

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/amanda/+bug/2099800/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to