[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bacula-devel] Regression 'make setup' broken in svn?


On Tue, Oct 07, 2008 at 06:41:13PM +0200, Kern Sibbald wrote:
> On Tuesday 07 October 2008 16:04:34 Graham Keeling wrote:
> > On Tue, Oct 07, 2008 at 02:31:13PM +0100, Graham Keeling wrote:
> > > I've just been trying to build the regression tests from svn (checked out
> > > last night).
> >
> > The reason that I want to do this is that I've been trying out the
> > VirtualFull feature again, and I think there are still a problem or two
> > with it. So, I'd like to prove it using the regression tests (or adding a
> > new test).
> >
> > Firstly, when running a VirtualFull, it doesn't seem to include all the
> > Incrementals since the last Full. So, for example, I might do...
> > a) Full
> > b) Incr
> > c) Incr
> > d) Incr
> > ...and then run a VirtualFull, and the log will only mention the backups
> > created in (a), (b) and (d), and miss out (c) completely. There doesn't
> > seem to be an obvious pattern to the ones that it misses out.
> 
> It only includes the ones that are needed (i.e. it optimizes).  If all the 
> files that are in backup c) are present in any future backup, then c) there 
> will be nothing to restore from c).
> 
> >
> > Secondly, the VirtualFull Job that I end up with has a StartTime the same
> > as the Full ('a' in the example above), which doesn't seem right.
> 
> It really should be setting it to the time of the last backup rather than the 
> last Full.  That sounds like a bug.  Can you verify?

I think so.

I added the line 'cp ${cwd}/tmp/log2.out /tmp' near the end of 
./tests/virtual-backup-test, so that I could read the output after the test.
There might be another way of doing that, but I don't know what it is.
Then I run ./tests/virtual-backup-test and look at /tmp/log2.out.
Near the end of the output, it does a 'list jobs', which follows (tidied up
a bit to fit nicely into the width of my console)...

+---+--------------+---------------------+---+---+----------+----------+
| J | Name         | StartTime           | T | L | JobFiles | JobBytes |
+---+--------------+---------------------+---+---+----------+----------+
| 1 | Vbackup      | 2008-10-07 17:11:19 | B | F | 4273     | 89581447 |
| 2 | Vbackup      | 2008-10-07 17:11:36 | B | I | 43       | 926899   |
| 3 | Vbackup      | 2008-10-07 17:11:39 | B | D | 86       | 3628723  |
| 4 | Vbackup      | 2008-10-07 17:11:44 | B | I | 44       | 2701853  |
| 6 | Vbackup      | 2008-10-07 17:11:44 | B | F | 4273     | 90224486 |
| 5 | RestoreFiles | 2008-10-07 17:11:47 | R | F | 4273     | 89581447 |
| 7 | RestoreFiles | 2008-10-07 17:12:17 | R | F | 4273     | 89581447 |
+---+--------------+---------------------+---+---+----------+----------+

You can see that JobId 6 has the same StartTime as JobId 4, which seems
reasonable, and is consistent with what you say.

However, I then change the Differential backup in ./tests/virtual-backup-test
to be an Incremental, and run the test again.
In this case, I would expect the 'list jobs' output to be very similar to the
above, only with an 'I' in the place of the 'D'.

So, here is what it actually looks like...


+---+--------------+---------------------+---+---+----------+----------+
| J | Name         | StartTime           | T | L | JobFiles | JobBytes |
+---+--------------+---------------------+---+---+----------+----------+
| 1 | Vbackup      | 2008-10-08 09:12:51 | B | F | 4273     | 89581447 |
| 2 | Vbackup      | 2008-10-08 09:13:18 | B | I | 43       | 926899   |
| 6 | Vbackup      | 2008-10-08 09:13:18 | B | F | 4273     | 90224486 |
| 3 | Vbackup      | 2008-10-08 09:13:23 | B | I | 43       | 2701824  |
| 4 | Vbackup      | 2008-10-08 09:13:27 | B | I | 44       | 2701853  |
| 5 | RestoreFiles | 2008-10-08 09:13:30 | R | F | 4273     | 89581447 |
| 7 | RestoreFiles | 2008-10-08 09:14:07 | R | F | 4273     | 89581447 |
+---+--------------+---------------------+---+---+----------+----------+

You can see that JobId 6 now has the same StartTime as JobId 2, which was the
first Incremental of the three.
So, I haven't shown that it has the same StartTime as the first Full, but it
definitely looks like there is something dodgy going on.


-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Bacula-devel mailing list
Bacula-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/bacula-devel


This mailing list archive is a service of Copilotco.