On Wednesday 09 May 2007 11:12, Mariusz Czulada wrote:
> Hi,
>
> Primo: I checked out the SVN repository with no problem.
> Will try to use it as a reference in the future.
>
> Secundo: Forgot to tell about configuration. This is an IBM
> xSeries host with CentOS 4.3. DDS-5 internal drive (working
> perfectly), and IBM TS3310 tape library, with two LTO-3
> drives connected directly (i mean without FC switch) to
> two QLogic QLA2460 HBA. IBM tape drivers taken from
> IBM site. Catalog is created in PostgreSQL 8.1.5.
> Bacula 2.0.2. Both pg and bacula compiled from
> sources.
Be very careful with the Fibre Channel. They have a tendency to cache writing
to the tape drive, which *must* be turned off for Bacula to work correctly.
Don't be content with just running the "test" command -- you should also run
the "fill" command, to verify that your FC is not caching the end of tape
status and creating bad volumes when they fill !!!!!
>
> Ad rem:
> I didn't give up with 2.0.2 yesterday. After btape failed
> in second test. The error I got was:
>
> ----------------------------
>
> btape: dev.c:2405 dev_get_os_pos
>
> ============================
>
> I took at the code. Near line 2405 (ver. 2.0.2) found:
>
> ----------------------------
>
> stat = tape_ioctl(fd, MTIOCTOP, (char *)&mt_com);
> if (stat < 0 || !dev_get_os_pos(this, &mt_stat)) {
> berrno be;
Yes, I saw that and believe that I have fixed it in 2.1.8.
>
> ============================
>
> So, I made a "bad" change:
>
> ----------------------------
>
> stat = tape_ioctl(fd, MTIOCTOP, (char *)&mt_com);
> ! dev_get_os_pos(this, &mt_stat);
> ! if (stat < 0) {
> berrno be;
>
> ============================
>
> Awful, but it works. Precisely, it still failed this test,
> but btape continued with other ones. And the rest of tests
> gave me hints how to correct my "Device" entries in
> SD config. I got:
>
> ----------------------------
>
> Hardware End of Medium = No
> Fast Forward Space File = No
> BSF at EOM = yes
When you are sure that this works, please send me your complete Device
resource reminding me of exactly what system and tape drive you have, and I
will add it to the default bacula-sd.conf file so that others may benefit
from your testing.
By the way, if the above directives apply to your DDS-3 tape, OK, but they do
not at all seem appropriate for an LTO-3 tape, which is a modern drive.
Something seems way less than optimal here. You might check FC caching (it
is on by default for HP manufactured FCs) and check how IBM sets up the
default drive parameters -- see the manual.
>
> ============================
>
> Now, original btape works correctly. So, my little suggestion
> about modifying btape util is to allow (query about) to
> continue the tests even if one of the tests fails. Hints
> are useful and help to correctly configure the drive...
> unless you see them :-)
>
>
> ------
>
> Summa:
>
> The bad news is I still cannot use my TS3310. The problem
> occurs when tape is reinserted into drive. Bacula cannot
> add another job to the tape. The volume is marked with
> "error" and the new one (from "Scratch" in my case)
> is used. Everything works fine, jobs are stored correctly,
> until newly used tape is removed from drive by changer.
If Bacula cannot add another file after writing the tape (after
removing/reinserting the tape, or stopping and restarting the SD) then 99%
likely it is a problem with tape drive vs Bacula not agreeing on how many EOF
marks to put at the end of a file.
>
> Will tell more/ask later, plan to do more tests to get
> more information which, I hope, could help you to help me :-)
Regards,
Kern
>
>
> Regards
>
> Mariusz
>
>
> Dnia 8-05-2007 o godz. 21:34 Kern Sibbald napisał(a):
> > Hello,
> >
> > One of your ioctl() calls seems to be broken or always returning
> > a bad status.
> >
> > Try using the btape in the trunk of the SVN. It has a number
> > of little fixes that could clarify your problem. If you don't
> > know how to use the SVN, the list can probably help you or
> > wait until 2.1.8 is released in BETA later this week.
> >
> >
> > >Date: Tue, 08 May 2007 09:30:56 +0200
> > >From: "Mariusz Czulada" <[EMAIL PROTECTED]>
> > >Subject: [Bacula-users] Odp: Another problem with TS3310
> > >To: "bacula-users" <bacula-users@lists.sourceforge.net>
> > >Message-ID: <[EMAIL PROTECTED]>
> > >Content-Type: text/plain; charset=iso-8859-2
> >
> > >Hi again,
> >
> > >I tried one more thing with btape, and this is the result:
> >
> > >==============================================
> >
> > >*rewind
> > >btape: btape.c:469 Rewound "Dev:TS3310-drv0" (/dev/IBMtape1n)
> > >*status
> > > Bacula status: file=0 block=0
> > > Device status: BOT ONLINE file=-1 block=0
> > >btape: btape.c:1799 Device status: 133. ERR=dev.c:1298 ioctl MTFSF error
> > >on "Dev:TS3310-drv0" (/dev/IBMtape1n). ERR=Success.
> >
> > >*fsf
> > >btape: dev.c:2405 dev_get_os_pos
> > >btape: btape.c:1497 Bad status from fsf. ERR=dev.c:1298 ioctl MTFSF
> > >error on "Dev:TS3310-drv0" (/dev/IBMtape1n). ERR=Success.
> >
> > >*status
> > > EOD Bacula status: file=0 block=0
> > > Device status: ONLINE file=-1 block=1001
> > >btape: btape.c:1799 Device status: 161. ERR=dev.c:1298 ioctl MTFSF error
> > >on "Dev:TS3310-drv0" (/dev/IBMtape1n). ERR=Success.
> >
> > >==========================
> >
> > >Can this be more helpful? This is a critical issue for me, because the
> > >main hardware component of this backup subsystem is currently useless.
> >
> > >Dnia 7-05-2007 o godz. 14:46 Mariusz Czulada napisa?(a):
> > >> Hi all,
> > >>
> > >> I have another, more serious problem with my IBM library. Here is an
> > >> output from btape:
> > >>
> > >> -----------------------------------------
> > >>
> > >> [EMAIL PROTECTED] sbin]# ../sbin/btape -c bacula-sd.conf -d 99
> > >> /dev/IBMtape1n
> > >> Tape block granularity is 1024 bytes.
> > >>
> > >> [...]
> > >>
> > >> btape: btape.c:368 open device "Dev:TS3310-drv0" (/dev/IBMtape1n): OK
> > >> *test
> > >> === Write, rewind, and re-read test ===
> > >>
> > >> [...]
> > >>
> > >> === Test Succeeded. End Write, rewind, and re-read test ===
> > >> === Write, rewind, and position test ===
> > >>
> > >> [...]
> > >> > btape: btape.c:962 Rewind OK.
> > >> Reposition to file:block 0:4
> > >> btape: dev.c:1477 fsr 4
> > >> Block 5 re-read correctly.
> > >> Reposition to file:block 0:200
> > >> btape: dev.c:1477 fsr 195
> > >> Block 201 re-read correctly.
> > >> Reposition to file:block 0:999
> > >> btape: dev.c:1477 fsr 798
> > >> Block 1000 re-read correctly.
> > >> Reposition to file:block 1:0
> > >> btape: dev.c:2405 dev_get_os_pos
> > >> btape: btape.c:1004 Reposition error.
> > >>
> > >> -----------------------------------------
> > >> [...]
> > >> -----------------------------------------
> > >>
> > >> [EMAIL PROTECTED] sbin]# mt -f /dev/IBMtape1n tell
> > >> At block 1001.
> > >>
> > >> -----------------------------------------
> > >>
> >
> > >Reagards,
> >
> > >Mariusz
>
> ----------------------------------------------------
> Ekstraliga żużlowa - sprawdź koniecznie kolejne rozgrywki! Kliknij:
> http://klik.wp.pl/?adr=www.ekstraliga.wp.pl&sid=1129
>
>
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
