On Fri, 2006-05-05 at 21:17 -0400, Peter Sjoberg wrote:
> On Fri, 2006-05-05 at 16:06 -0400, Ryan Novosielski wrote:
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> >
> > It absolutely does work, at least with single drives. I do this
> > currently (writing larger fulls to a DAT72 drive and smaller
> > incrementals to a DDS4).
> >
> > Here's an example:
> >
> > Schedule {
> > Name = "UMD-F13T-Inc"
> > Run = Level=Full Storage=helios_DAT72 1st,3rd tue at 21:00
> > Run = Level=Incremental 1st,3rd mon,wed-fri at 23:00
> > Run = Level=Incremental 2nd,4th-5th mon-fri at 23:00
> > }
>
> I was looking in different places but I guess Job and Pool was the wrong
> place.
> I will try this and see how it works out but after reading the doc it
> seems like a good change that it will work.
>
> Based on Kerns' comment I guess it's some bug related to multiple
> autochangers that's fixed with the latest release so I will also upgrade
> to 1.38.9 before giving it a try.
Tried it out and the bad news is that it doesn't work for several
reasons, some minor and some major and in the end I think it's some work
left to do and bugs to fix in this area before I can really use it.
Setup:
Schedule {
Name = "MonthlyCycle"
Run = Full Storage=ADIC 2nd sun at 0:35
Run = Differential Storage=TL892 1st,3rd-5th sun at 3:05
Run = Incremental Storage=TL892 mon-sat at 3:05
}
in ADIC I loaded 22 tapes from the Full pool
in TL892 I loaded a mix of DIFF and INC tapes
I did update slots on both autochangers
Directly I then issued "run Backup_janeway Level=Full yes" and followed
what was going on.
It took a look on TL892 and then loaded a tape, this when I expected it
to load it in ADIC. Of course this strange behavior can be expected
since the storage is defined in the scedule and not the pool, it doesn't
really know the connection and just use the default job storage which
happens to be TL892. I guess it could have know that all the "FUL" tapes
where in the other loader by looking in the database record but I guess
the default job storage takes precedence.
I then issued "run Backup_lister Level=Full Storage=ADIC yes" and this
time it loaded a tape in ADIC (same slot as it did on load earlier
TL892) and completed a full backup.
I now realized that if an incr/diff is upgraded to Full it will still
try to use the wrong autochanger so I would have to be careful in that
area but this would still not be a show stopper for me so I left it to
run a normal incremental over night.
In the morning I find that all jobs failed.
"Backup_janeway" starts of with "Prior failed job found. Upgrading to
Full". Since the test before failed it's upgraded to full and my
suspicions are confirmed, it still try to use TL892. The rest of that
job error is
06-May 03:05 sisko-dir: Prior failed job found. Upgrading to Full.
06-May 03:05 sisko-dir: Start Backup JobId 8181,
Job=Backup_janeway.2006-05-06_03.05.00
06-May 03:05 sisko-sd: 3301 Issuing autochanger "loaded drive 0" command.
06-May 03:05 sisko-sd: 3302 Autochanger "loaded drive 0", result: nothing
loaded.
06-May 03:05 sisko-sd: 3304 Issuing autochanger "load slot 3, drive 0" command.
06-May 03:09 sisko-sd: 3305 Autochanger "load slot 3, drive 0", status is OK.
06-May 03:09 sisko-sd: 3301 Issuing autochanger "loaded drive 0" command.
06-May 03:09 sisko-sd: 3302 Autochanger "loaded drive 0", result is Slot 3.
06-May 03:14 sisko-sd: Backup_janeway.2006-05-06_03.05.00 Fatal error:
dev.c:362 dev.c:356 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0):
ERR=Device or resource busy
06-May 03:14 sisko-sd: Backup_janeway.2006-05-06_03.05.00 Fatal error:
device.c:296 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0):
ERR=dev.c:356 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0): ERR=Device
or resource busy
06-May 03:14 janeway-fd: Backup_janeway.2006-05-06_03.05.00 Fatal error:
job.c:1602 Bad response to Append Data command. Wanted 3000 OK data
, got 3903 Error append data
So it fails some issue around the device being busy. After all, I had
several jobs kicked off at the same time and it seems like they all
failed to open the device.
06-May 03:14 sisko-dir: Start Backup JobId 8185,
Job=Backup_lister.2006-05-06_03.05.04
06-May 03:34 sisko-sd: 3301 Issuing autochanger "loaded drive 0" command.
06-May 03:34 sisko-sd: 3302 Autochanger "loaded drive 0", result is Slot 6.
06-May 03:34 sisko-sd: 3301 Issuing autochanger "loaded drive 0" command.
06-May 03:34 sisko-sd: 3302 Autochanger "loaded drive 0", result is Slot 6.
06-May 03:39 sisko-sd: Backup_lister.2006-05-06_03.05.04 Fatal error: dev.c:362
dev.c:356 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0): ERR=Device or
resource busy
06-May 03:39 sisko-sd: Backup_lister.2006-05-06_03.05.04 Fatal error:
device.c:296 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0):
ERR=dev.c:356 Unable to open device "TL0" (/dev/scsi/nsth5-1c0i4l0): ERR=Device
or resource busy
06-May 03:39 lister-fd: Backup_lister.2006-05-06_03.05.04 Fatal error:
job.c:1602 Bad response to Append Data command. Wanted 3000 OK data
, got 3903 Error append data
At the moment I thing going back to only using one tapeloader is the
only thing I can do.
>
> Thanks
>
> /ps
> >
> > I'm guessing I don't have Storage labeled for incrementals because the
> > other drive is the default.
> >
> > ---- _ _ _ _ ___ _ _ _
> > |Y#| | | |\/| | \ |\ | | |Ryan Novosielski - User Support Spec. III
> > |$&| |__| | | |__/ | \| _| |[EMAIL PROTECTED] - 973/972.0922 (2-0922)
> > \__/ Univ. of Med. and Dent.|IST/AST - NJMS Medical Science Bldg - C630
> >
> >
> > Peter Sjoberg wrote:
> > > I'm asking for suggestions on how I can best utilize two tape libraries.
> > > One tape library is a 10 x DLT7000
> > > The other is 15 x DLT8000
> > >
> > > At first I was thinking of putting the FULL backup to the bigger library
> > > and run diff and inc to the smaller one but I don't know the best way to
> > > do that.
> > > I kind of expected to be able to do that in one single job but that
> > > doesn't seem to be possible to have one storage for full and another for
> > > diff/inc.
> > > Setting up separate jobs won't work since then the diff/inc job won't
> > > find the "Full" to base it on.
> > >
> > > Any suggestions ?
> > >
> > >
> > > In case it matters, I'm running Bacula 1.38.5
> > >
> > >
> > >
> > -----BEGIN PGP SIGNATURE-----
> > Version: GnuPG v1.4.2.2 (MingW32)
> >
> > iD8DBQFEW7A7mb+gadEcsb4RAloZAKCR76mOhI55amkkfyo1/7HKa5EMMACeNL1+
> > 3VpBgdhZwpllffJns/5oGJQ=
> > =9SFj
> > -----END PGP SIGNATURE-----
>
>
> -------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
