Hi,
07.11.2007 15:37,, Win Htin wrote::
> Hi Arno,
>
...
> This brings up a question. Am I correct to assume bacula actually writes
> the volume (barcode label in my case) and the associated pool's info
> onto the tape?
Yes, I think so. Try bls to look at the tape contents, for example.
> Also, I encountered a different error when I started running automated
> backups last evening.
>
> Scenario:
> Number of jobs : 5 ( I have 5 RH EL4 servers including the Bacula tape
> back server)
> Schedule : 1
> Schedule {
> Name = "Daily"
> Run = Full 4th-5th sun at 20:00
> Run = Differential mon-sat at 23:00
> Run = Differential 1st-3rd sun at 23:00
> }
>
> Jobdefinition :1
> JobDefs {
> Name = "BACKUPS"
> Type = Backup
> FileSet = "UnixSet"
> Schedule = "Daily"
> Storage = TS3200_1
> SpoolData = yes
> Messages = Standard
> Pool = AllFulls
> Full Backup Pool = AllFulls
> Incremental Backup Pool = AllDifferentials
> Differential Backup Pool = AllDifferentials
> Priority = 10
> }
>
> When the backups kicked in at 11pm, volume 400AAL from AllDifferentials
> pool was in the tape drive and the first job that started was a
> differential backup of the server nfsserver, which found the correct
> tape in the tape drive. (The full backup was performed earlier on this
> server when I ran the backups manually.)
> 06-Nov 23:00 nfsserver-dir JobId 4: Start Backup JobId 4,
> Job=NFSSERVER.2007-11-06_23.00.08
> 06-Nov 23:00 nfsserver-dir JobId 4: Using Device "LTO4_1"
> 06-Nov 23:00 nfsserver-sd JobId 4: Spooling data ...
> ...
> ...
> 06-Nov 23:00 nfsserver-sd JobId 4: Committing spooled data to Volume
> "400AAL". Despooling 267,152,653 bytes ...
> 06-Nov 23:00 nfsserver-sd JobId 4: Despooling elapsed time = 00:00:03,
> Transfer rate = 89.05 M bytes/second
> ...
> Build OS: x86_64-unknown-linux-gnu redhat Enterprise release
> JobId: 4
> Job: NFSSERVER.2007-11-06_23.00.08
> Backup Level: Differential, since=2007-11-06 15:50:38
> Client: "nfsserver-fd" 2.2.5 (09Oct07)
> x86_64-unknown-linux-gnu,redhat,Enterprise release
> FileSet: "UnixSet" 2007-11-06 13:51:41
> Pool: "AllDifferentials" (From Job DiffPool override)
> Storage: "TS3200_1" (From Pool resource)
> Scheduled time: 06-Nov-2007 23:00:00
> Start time: 06-Nov-2007 23:00:00
> End time: 06-Nov-2007 23:00:12
Note that this job was finished at 23:00:12.
> Elapsed time: 12 secs
> Priority: 10
> FD Files Written: 49
> SD Files Written: 49
> FD Bytes Written: 266,897,681 (266.8 MB)
> SD Bytes Written: 266,903,369 ( 266.9 MB)
> Rate: 22241.5 KB/s
> Software Compression: None
> VSS: no
> Encryption: no
> Volume name(s): 400AAL
> Volume Session Id: 3
> Volume Session Time: 1194380622
> Last Volume Bytes: 513,321,984 (513.3 MB)
> Non-fatal FD errors: 0
> SD Errors: 0
> FD termination status: OK
> SD termination status: OK
> Termination: Backup OK
> The backup went OK without a hitch.
>
> At the same time, the other client jobs kicked in and being a brandnew
> backup, those jobs all started as Full backups.
>
> 06-Nov 23:00 nfsserver-dir JobId 5: sql_find.c:134 No Job record found: ERR=
> CMD=SELECT StartTime FROM Job WHERE JobStatus='T' AND Type='B' AND
> Level='F' AND Name='RH006' AND ClientId=2 AND FileSetId=1 ORDER BY
> StartTime DESC LIMIT 1
> 06-Nov 23:00 nfsserver-dir JobId 5: No prior or suitable Full backup
> found in catalog. Doing FULL backup.
> 06-Nov 23:00 nfsserver-dir JobId 5: Start Backup JobId 5,
> Job=RH006.2007-11-06_23.00.09
> 06-Nov 23:00 nfsserver-dir JobId 5: Using Device "LTO4_1"
> 06-Nov 23:00 nfsserver-sd JobId 5: 3307 Issuing autochanger "unload slot
> 1, drive 0" command.
> 06-Nov 23:01 nfsserver-sd JobId 5: 3304 Issuing autochanger "load slot
> 22, drive 0" command.
> 06-Nov 23:01 nfsserver-sd JobId 5: 3305 Autochanger "load slot 22, drive
> 0", status is OK.
Not that the tape change was finished after the other job was
terminated. It probably started after that fisrt job, bit the output
timestamps don't show the seconds :-)
> 06-Nov 23:01 nfsserver-sd JobId 5: 3301 Issuing autochanger "loaded?
> drive 0" command.
> 06-Nov 23:01 nfsserver-sd JobId 5: 3302 Autochanger "loaded? drive 0",
> result is Slot 22.
> 06-Nov 23:01 nfsserver-sd JobId 5: Volume "626AAL" previously written,
> moving to end of data.
> 06-Nov 23:02 nfsserver-sd JobId 5: Error: Unable to position to end of
> data on device "LTO4_1" (/dev/IBMtape0): ERR=dev.c:1326 read error on
> "LTO4_1" (/dev/IBMtape0). ERR=Input/output error.
Check what's in the catalog regarding this tape, and what's actually
on it.
use 'llist volume=626AAL' in bconsole, for example. I suspect that
there are some of Baculas files (and bytes, accordingly) on it, while
the tape actually only contains one or two EOFs.
...
> This Full backup Job was also successful BUT it put the volume 626AAL
> into error. The following full backups for the remaining clients all
> went ok on volume 625AAL. By reconstructing the event both differential
> and full backups started at the same time and the differential job took
> over the tape drive
If that happened that would be a mjor bug in Bacula, preventing it
from working in many setups that are known to work :-)
> making the full backup to think tape 626AAL is faulty.
> ( Reading volume 400AAL thinking it was 626AAL and not finding eof? )
>
> Please help how I should configure my jobs so that this sort of tape
> contention for different pools will not occur. Thanks in advance.
First ensure the tape status for 626AAL corresponds to what is on
tape. If it doesn't, that's the problem. If it does, you might simply
delete that volume from Baculas catalog, wipe the tape, and label it
again. If the problem persists, the tape might be damaged.
Arno
> Regards,
>
> Win
>
>
>
> ------------------------------
>
> Message: 8
> Date: Tue, 06 Nov 2007 21:05:22 +0100
> From: Arno Lehmann <[EMAIL PROTECTED] <mailto:[EMAIL PROTECTED]>>
> Subject: Re: [Bacula-users] Bacula backup running amok with "not
> configured to autolabel Volumes" messages
> To: bacula-users@lists.sourceforge.net
> <mailto:bacula-users@lists.sourceforge.net>
> Message-ID: < [EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Hi,
>
> 06.11.2007 20:36,, Win Htin wrote::
> > Hi folks,
> >
> > I have a RedHat EL4 server connected through FC to an IBM TS3200
> tape
> > changer with single LTO4 drive with 44 slots. The bacula version is
> > 2.2.5. I have set the label to use the barcode labels and wiped
> all 25
> > tapes in the changer clean with the command "mt rewind && mt weof
> && mt
> > rewind" to make sure all previous label and data entries from earlier
> > tests are completely gone.
> >
> > Entry from bacula-sd.conf:
> >
> > Label Type = IBM
> > Check Labels = yes
> > LabelMedia = no;
> >
> > Following is the sequence of my test run:
> >
> > 1. Checked status of the autochanger just before the job was started.
> > Confirmed NO tapes loaded in the tape drive.
> > 2. *list media
> ... there are volumes available...
>
> > 3. *run NFSSERVER
> > 06-Nov 13:51 nfsserver-dir JobId 1: No prior Full backup Job
> record found.
> > 06-Nov 13:51 nfsserver-dir JobId 1: No prior or suitable Full backup
> > found in catalog. Doing FULL backup.
> > 06-Nov 13:51 nfsserver-dir JobId 1: Start Backup JobId 1,
> > Job=NFSSERVER.2007-11-06_13.51.03
> > 06-Nov 13:51 nfsserver-dir JobId 1: Using Device "LTO4_1"
> > 06-Nov 13:51 nfsserver-sd JobId 1: 3301 Issuing autochanger "loaded?
> > drive 0" command.
> > 06-Nov 13:51 nfsserver-sd JobId 1: 3302 Autochanger "loaded?
> drive 0",
> > result: nothing loaded.
> > 06-Nov 13:51 nfsserver-sd JobId 1: 3304 Issuing autochanger "load
> slot
> > 21, drive 0" command.
> > 06-Nov 13:52 nfsserver-sd JobId 1: 3305 Autochanger "load slot
> 21, drive
> > 0", status is OK.
> > 06-Nov 13:52 nfsserver-sd JobId 1: 3301 Issuing autochanger "loaded?
> > drive 0" command.
> > 06-Nov 13:52 nfsserver-sd JobId 1: 3302 Autochanger "loaded?
> drive 0",
> > result is Slot 21.
> > 06-Nov 13:52 nfsserver-sd JobId 1: Warning: Device "LTO4_1"
> > (/dev/IBMtape0) not configured to autolabel Volumes.
>
> Ignore the warning for now, it's not an error.
>
> The problem is that this volume is not recognized.
>
> I think you should do the following:
>
> 1. umount the device from bconsole
> 2. issue 'setdebug level=400 trace=1 sd=LTO4_1',
> 3. wait for any messages, then 'setdebug level=0 trace=0 sd=LTO4_1'
> 4. find the lines in the resulting nfsserver-sd.trace file in the
> SD's working directory where the tape is not accessible (you might
> see something regarding a "rewind error".
> 5. use mtx to unload the tape from the drive
> 6. use mtx to load that same tape to the drive again (don't ask...)
> 7. in bconsole, issue 'mount sd=LTO4_1 slot=21'
>
> According to my theory, your job should run then.
> If it does, report back, please - it seems as this is a problem that
> affects several installations.
>
> If I'm guessing right, you'll find that everything works ok as long as
> tapes from among one pool are exchanged, but once you use another
> pool, Bacula doesn't accept the tape because it can't rewind it.
>
> For me, this looks really weird.
>
> If I get this confirmed by others, I'll file a bug report - unless
> someone beats me at it :-)
>
> Arno
>
> --
> Arno Lehmann
> IT-Service Lehmann
> www.its-lehmann.de <http://www.its-lehmann.de>
> ------------------------------
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems? Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Bacula-users mailing list
> Bacula-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/bacula-users
--
Arno Lehmann
IT-Service Lehmann
www.its-lehmann.de
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
