On 11/04/11 08:14, Martin Simmons wrote:
> The volumes record the Client and FileSet names, not their ids, so it can't
> fail like that. However, as Kern said, there will be problems with jobs
> spanning volumes if you don't scan them all at once and in the correct order.
> Bscan could probably be smarter about this.
OK, the theorem is wrong then. However, even usign a BSR file to scan
all the 5.0.3 volumes in the correct order in a single bscan session,
this is my result:
mysql> SELECT Client.Name Client, FileSet.FileSet, count(Job.JobId) FROM
Job JOIN FileSet ON Job.FileSetId = FileSet.FileSetId JOIN Client ON
Job.ClientId = Client.ClientId GROUP BY Job.ClientId, Job.FileSetId;
+----------------+-------------------+------------------+
| Client | FileSet | count(Job.JobId) |
+----------------+-------------------+------------------+
| babylon5 | Babylon4 Full Set | 2 |
| babylon5 | OpenBSD Full Set | 36 |
| babylon5 | Gentoo Full Set | 5 |
| spikyfishthing | Babylon4 Full Set | 2 |
| spikyfishthing | OpenBSD Full Set | 40 |
| babylon4 | Babylon4 Full Set | 5 |
| babylon4 | Catalog Snapshot | 41 |
| babylon4 | OpenBSD Full Set | 27 |
| babylon4 | Gentoo Full Set | 9 |
+----------------+-------------------+------------------+
9 rows in set (0.08 sec)
mysql> SELECT Job.Name Job, FileSet.FileSet, count(Job.JobId) FROM Job
JOIN FileSet ON Job.FileSetId = FileSet.FileSetId GROUP BY Job.Name,
Job.FileSetId;
+-----------------------+-------------------+------------------+
| Job | FileSet | count(Job.JobId) |
+-----------------------+-------------------+------------------+
| Babylon4 Backup | Babylon4 Full Set | 5 |
| Babylon4 Backup | OpenBSD Full Set | 27 |
| Babylon4 Backup | Gentoo Full Set | 9 |
| Babylon5 Backup | Babylon4 Full Set | 2 |
| Babylon5 Backup | OpenBSD Full Set | 36 |
| Babylon5 Backup | Gentoo Full Set | 5 |
| MySQL Backup | Catalog Snapshot | 41 |
| Spikyfishthing Backup | Babylon4 Full Set | 2 |
| Spikyfishthing Backup | OpenBSD Full Set | 40 |
+-----------------------+-------------------+------------------+
9 rows in set (0.01 sec)
Whatever's happening, it looks like an empty catalog cannot be safely
(which is to say, usably) restored by using 5.2.1's bscan to scan all
the volumes.
Interestingly, all of the Catalog Snapshot jobs are correct. However,
that seems to be the ONLY job that's consistently got the correct
Fileset recorded in the bscanned Catalog.
(The MySQL Backup job and the Babylon4 Backup job both run on client
babylon4, which is also the Director.)
--
Phil Stracchino, CDK#2 DoD#299792458 ICBM: 43.5607, -71.355
ala...@caerllewys.net ala...@metrocast.net p...@co.ordinate.org
Renaissance Man, Unix ronin, Perl hacker, SQL wrangler, Free Stater
It's not the years, it's the mileage.
------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Bacula-devel mailing list
Bacula-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-devel
