Re: core file generated during restore

[John R. Jackson]

What program generated the core file?  Run "file core" and it should
tell you.

Once you know that, run gdb or dbx on the program that failed and the
core file, then type "where" at the prompt and post those results.

John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]

RE: Amanda, FreeBSD 4.3-STABLE, and HP SureStore 12000E 4mm DAT changer--HELP

[Noonan, Sean P.]

Hi Thomas,

Thanks for responding!

I'll look into the 2.5.0 development version of amanda and compiling it
myself instead of using the ports tree.  Seems to me, though, that there
were no shapshots available, and that I had to use CVS.  Since I didn't
[don't] know how to use CVS, I skipped over this option.  Guess it's time to
learn CVS...

In term of your question regarding whether the tape drive and robot use
different SCSI id's, I'm not sure how to answer (I constatntly confuse SCSI
ids with LUNs, etc.).  Hopefully this answers you question:

su-2.05# camcontrol devlist   
 at scbus0 target 0 lun 0 (pass0,da0)
at scbus0 target 1 lun 0 (pass1,da1)
   at scbus0 target 3 lun 0 (pass2,sa0)
   at scbus0 target 3 lun 1 (pass3,ch0)

What I *think* this is telling me is that the tape and robot use the same
SCSI ID but different LUNs. (?).  Does this help?

Here's some more info, in case it helps you help me :0)

su-2.05# chio status
slot 0: 
slot 1: 
slot 2: 
slot 3: 
slot 4: 
slot 5: 
drive 0: 


su-2.05# chio params
/dev/ch0: 6 slots, 1 drive, 0 picker
/dev/ch0: current picker: 0


su-2.05# /usr/local/sbin/mtx -f /dev/pass2 inquiry
Product Type: Tape Drive
Vendor ID: 'HP  '
Product ID: 'C1553A  '
Revision: '9503'
Attached Changer: No


su-2.05# /usr/local/sbin/mtx -f /dev/pass2 next   
Too many Data Transfer Elements Reported


 -Original Message-
From:   Thomas Hepper [mailto:[EMAIL PROTECTED]] 
Sent:   Thursday, May 31, 2001 12:09 PM
To: Sean Noonan
Cc: [EMAIL PROTECTED]
Subject:Re: Amanda, FreeBSD 4.3-STABLE, and HP SureStore 12000E 4mm
DAT changer--HELP

Hi,
On Thu, May 31, 2001 at 08:27:41AM -0700, Sean Noonan wrote:
> 4. My biggest problem thus far is finding a "changer glue script" that
> works with the HP SureStore 12000E.  I've tried chg-chio, chg-mtx,
> chg-manual, chg-everthing.  The only thing I haven't tried, and my guess
> is that it would work, is chg-scsi.  Apparently it's broke since FreeBSD
> switched to CAM?  I notice the Makefile for the port has references to
> chg-scsi remarked out.  I searched the archives and found conflicting
> reports of chg-scsi working with FreeBSD.  One reference said chg-scsi has
> been broke since the introduction of CAM.  Another contradicted that one
> and said that the author had since fixed the problem.  Don't believe this
> to be the case, since chg-scsi didn't build automatically.  Tried manually
> making chg-scsi, but got numerous errors.  Can post errors to the list if
> anyone wants to see them.

The 2.4.2p2 version of chg-scsi is not the best for FreeBSD, can you
try to download amanda-2.5.0 (the development version) and try chg-scsi
from there (compile amanda, but don't install it, copy only chg-scsi)

The 2.5.0 version of chg-scsi has an option to create an sample config
for chg-scsi, try chg-scsi -genconf.
> 
> I guess the best thing that could happen is for someone to post or send me
> their WORKING configuration of FreeBSD 4.3, HP SureStore 12000E, and
> Amanda 2.4.2-p2.  Anybody got this combo out there???

Is this an device with different SCSI id's for the tape and the robot ?


Thomas

-- 
  ---
  |  Thomas Hepper[EMAIL PROTECTED] |
  | ( If the above address fail try   ) |
  | ( [EMAIL PROTECTED])|
  ---

RE: changing volumes on pioe input

[Mair, Wolfgang]

Hello,

I replying to share with you, what I found out with this problem. Thanks to
the one who replyed to me. Well, it took a while for me to find out the
following:

With solaris 8 you can have a mount opption "logging" (See the mount_ufs man
page). With this I get the error described below. With out this option the
restore works fine.
On the amanda page there ist a patch available for the advfs but I didn't
have the time to try it out if this would fix the problem. The description
doesn't really say something about my problem. Maybe somebody tried it
already.

All in all, what I can say now, be carefully with the logging option for
mounting on solaris 8!!

Wolfgang

-Original Message-
From: Mair, Wolfgang [mailto:[EMAIL PROTECTED]]
Sent: Montag, 14. Mai 2001 12:34
To: [EMAIL PROTECTED]
Subject: changing volumes on pioe input


Hello list,

can somebody please help me with the following problem:

I'm using amanda 2.4.2p2 on Solaris 8 and I don't have a tape device,
therfore I need to restore from the holdingdisk. The backup runs fine, I'm
doing a full backup (always-full) on each run. But with the restore I'm
having problems.

I'm trying to restore from the holding disk, after I mounted the directory
from the backup server. I use the following command:

dd if=/a/20010515/volvo._dev_dsk-c0t0d0s7.0 bs=32k skip=1 | ufsrestore -ifv
-

After entering the command, the ufsrestore runs fine for a while. Then I get
the following message:

42267+0 records in
42267+0 records out
changing volumes on pipe input
abort? [yn]

The only thing I can do at this point is to abort the restore. When I abort
it the system core is dumped. It looks to me as it only occurs while
restoring partitions arround or greater 1 Gbyte.
When I dump the image into a file f. e.:

dd if=volvo._dev_dsk-c0t0d0s7.0 bs=32k skip=1 of=volvopart7

and run ufsrestore:

ufsrestore -ifv ./volvopart7

ufsrestore expects a second volume. But there is no second volume...:-((( 


Any help or ideas would be appreciated


Wolfgang