bconsole's cancel doesn't remove job from sd

[Robert N]

Hi,

I'm running Bareos 15.2.3 ( Dir, SD, Console ) auf Centos6, tried to run a job ( Windows 15.2.2 Client ) with a ClientBeforeJob which didn't return ever, so the job got stuck.
I then did a cancel jobid=4546 in the bconsole
The job disappeared from status dir but it didn't disappear from status storage.

logs:

19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Fatal error: Socket error on RunBeforeNow command: ERR=Interrupted system call
19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Fatal error: Client "muc1-hyper-3" RunScript failed.
19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Fatal error: Network error with FD during Backup: ERR=Interrupted system call
19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Error: Director's comm line to SD dropped.
19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Fatal error: No Job status returned from FD.
19-Oct 12:51 muc1pro-backup-1-dir JobId 4546: Bareos muc1pro-backup-1-dir 15.2.3 (07Mar16):
Build OS: x86_64-redhat-linux-gnu redhat CentOS release 6.7 (Final)
JobId: 4546
Job: muc1-exchg.2016-10-19_11.38.33_11
Backup Level: Full
Client: "muc1-hyper-3" 15.2.2 (16Nov15) Microsoft Windows Server 2012 Datacenter Edition (build 9200), 64-bit,Cross-compile,Win64
FileSet: "muc1-exchg" 2016-07-04 15:51:40
Pool: "DiskWeekly" (From command line)
Catalog: "MyCatalog" (From Job resource)
Storage: "DiskStorageFull" (From Pool resource)
Scheduled time: 19-Oct-2016 11:38:16
Start time: 19-Oct-2016 11:38:35
End time: 19-Oct-2016 12:51:26
Elapsed time: 1 hour 12 mins 51 secs
Priority: 10
FD Files Written: 0
SD Files Written: 0
FD Bytes Written: 0 (0 B)
SD Bytes Written: 0 (0 B)
Rate: 0.0 KB/s
Software Compression: None
VSS: no
Encryption: no
Accurate: no
Volume name(s):
Volume Session Id: 380
Volume Session Time: 1475762430
Last Volume Bytes: 0 (0 B)
Non-fatal FD errors: 2
SD Errors: 0
FD termination status: Error
SD termination status: Error
Termination: Backup Canceled

The storage status in bconsole looks like:
Running Jobs:
Writing: Full Backup job muc1-exchg JobId=4546 Volume="Full-1333"
pool="DiskWeekly" device="DiskDeviceFull" (/data/bareos-storage/Full)
spooling=0 despooling=0 despool_wait=0
Files=0 Bytes=0 AveBytes/sec=0 LastBytes/sec=0
FDReadSeqNo=5 in_msg=4 out_msg=3 fd=5

Is that the expected behavior of cancel? or should it remove the jobs from the storage daemon as well?

Thank you.
Robert

[Frank Bergkemper]

Hi Robert,

it is the behaviour to expect.

As your comm line to the SD dropped you need to cancel the job
explicitly on the SD in that case.

cancel storage=<storage-name> jobid=<jobid>

Cheers
Frank