Hi again,

I looked at some of the files in our mystery differential backup (see last
post) again. I am pretty sure that some of the backed-up files are really old
and have not been modified since the last full backup (last Thursday, 

Here is one of these ancient files on disk:

[root@gundabad ~]# stat /home/stud/mx/simsel/csrc.rtl/3cuS_1_l.dat
  File: ‘/home/stud/mx/simsel/csrc.rtl/3cuS_1_l.dat’
  Size: 37              Blocks: 16         IO Block: 65536  regular file
Device: 2eh/46d Inode: 2204123     Links: 1
Access: (0644/-rw-r--r--)  Uid: ( 1033/      mb)   Gid: ( 1001/    stud)
Context: system_u:object_r:nfs_t:s0
Access: 2016-10-09 04:19:46.000000000 +0200
Modify: 2010-12-03 09:37:36.000000000 +0100
Change: 2016-09-06 18:32:55.000010000 +0200
 Birth: -

With these stats, I would assume that the file would not be part of the
differential, since they were also part of the full (job id 2152):

* list files jobid=2152
| /home/stud/mx/simsel/csrc.rtl/3cuS_1_l.dat

We have restored the file from the full backup, and it is identical to the
file which ended up again in the differential.

I am now a bit concerned about the integrity of all of our backups. We did
have problems on the _previous_ full Backup that the partition holding the
Postgres database had overflowed (due to the database exceeding 13GB). We have
since fixed that (increasing the size of the file system). The previous
differential on Sunday (2016-10-9) also completed without any errors or
warnings, so I had assumed that database problems (if any) due to the overflow
had been resolved.

So, any ideas what is happening here? Is there an SQL command to find out what
Bacula thinks about these files (why they are considered to be different, thus
ending up in differential backup)?


Attachment: signature.asc
Description: OpenPGP digital signature

Check out the vibrant tech community on one of the world's most 
engaging tech sites, SlashDot.org! http://sdm.link/slashdot
Bacula-users mailing list

Reply via email to