I reported this before, but never got a response:
Bacula 5.2.12 on CentOS 6, HP 1x8 LTO-3 autoloader (no barcode
reader), doing daily incremental backups to pool "Daily" and
full backups to pool "Weekly" or "Monthly" on weekends.
On Monday mornings after a "Monthly" backup, the "Monthly" tape
is manually replaced by the one for the next month, followed by
bconsole command "update LTO-3 slots=8 scan".
The same evening, when the next "Daily" backup run starts,
Bacula emits a spurious notification mail:
--------
Subject: Bacula: Intervention needed for backup.2013-02-04_20.05.00_36
Date: Mon, 04 Feb 2013 20:05:12 +0100 (CET)
From: (Bacula) <backup AT phoenixsoftware DOT de>
To: backup AT phoenixsoftware DOT de
04-Feb 20:05 bacula-sd JobId 6198: Job backup.2013-02-04_20.05.00_36 is
waiting. Cannot find any appendable volumes.
Please use the "label" command to create a new Volume for:
Storage: "LTO-3" (/dev/nst0)
Pool: Daily
Media type: LTO-3
------
After that, it proceeds to needlessly recycle a volume from the
"Daily" pool, only to load, five minutes later, the tape it should
have used in the first place and which had been sitting in the changer
in status "Append" all the time. Excerpt from the log, sorted roughly
chronologically (unfortunately the timestamps only have resolution
to the minute):
04-Feb 20:05 backup-dir JobId 6198: Start Backup JobId 6198,
Job=backup.2013-02-04_20.05.00_36
04-Feb 20:05 backup-dir JobId 6198: Using Device "LTO-3"
04-Feb 20:05 backup-dir JobId 6199: Start Backup JobId 6199,
Job=bombur.2013-02-04_20.05.00_37
04-Feb 20:05 backup-dir JobId 6199: There are no more Jobs associated
with Volume "Daily_0". Marking it purged.
04-Feb 20:05 backup-dir JobId 6199: All records pruned from Volume
"Daily_0"; marking it "Purged"
04-Feb 20:05 backup-dir JobId 6199: Recycled volume "Daily_0"
04-Feb 20:05 backup-dir JobId 6199: Using Device "LTO-3"
04-Feb 20:05 bacula-sd JobId 6198: Job backup.2013-02-04_20.05.00_36 is
waiting. Cannot find any appendable volumes.
04-Feb 20:10 bacula-sd JobId 6198: 3307 Issuing autochanger "unload slot
8, drive 0" command.
04-Feb 20:10 bacula-sd JobId 6198: 3304 Issuing autochanger "load slot
2, drive 0" command.
04-Feb 20:11 bacula-sd JobId 6198: 3305 Autochanger "load slot 2, drive
0", status is OK.
04-Feb 20:11 bacula-sd JobId 6198: Volume "Daily_1" previously written,
moving to end of data.
State of pool "Daily" *after* that run:
+---------+------------+-----------+---------+-----------------+----------+--------------+---------+------+-----------+-----------+---------------------+
| MediaId | VolumeName | VolStatus | Enabled | VolBytes |
VolFiles | VolRetention | Recycle | Slot | InChanger | MediaType |
LastWritten |
+---------+------------+-----------+---------+-----------------+----------+--------------+---------+------+-----------+-----------+---------------------+
| 18 | Daily_0 | Recycle | 1 | 1 |
0 | 518,400 | 1 | 1 | 1 | LTO-3 | 2013-01-09
02:05:17 |
| 34 | Daily_1 | Append | 1 | 219,879,346,176 |
65 | 518,400 | 1 | 2 | 1 | LTO-3 | 2013-02-04
23:19:02 |
| 35 | Daily_2 | Full | 1 | 799,806,099,456 |
320 | 518,400 | 1 | 3 | 1 | LTO-3 | 2013-01-30
20:18:11 |
+---------+------------+-----------+---------+-----------------+----------+--------------+---------+------+-----------+-----------+---------------------+
Note how all three volumes are InChanger and how volume Daily_0,
last written on 2013-01-09, has been recycled even though volume
Daily_1 is in Append state and still has lots of room.
Why? Any ideas?
TIA
T.
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------------
Free Next-Gen Firewall Hardware Offer
Buy your Sophos next-gen firewall before the end March 2013
and get the hardware for free! Learn more.
http://p.sf.net/sfu/sophos-d2d-feb _______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users
|
