Conference decwet::networker

547.0. "Bad tape and NSR stopped with a DR Watson file" by NQOS01::alfras1_port38.alf.dec.com::WORKBENCH () Wed Apr 02 1997 14:02

I have a Alpha 4100 running NT 3.51 with Service Pack 5 with NSR 4.3 and a 
TL812 Autochanger.

The other night during a backup, \\.\Tape0(J) loaded a tape and I guess the 
tape was bad so the 'c:\win32app\nsr\log\message.' file showed tape errors 
and an event log was posted, NSR stopped and a Dr Watson log file was 
written.  What files do I need to save so someone can look at this problem?

The portion of the message file that showed until NSR stopped:

Mar 31 03:28:01 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (warning) 
\\.\Tape0 moving: fsr 759: A tape access reached a filemark. (1101)
Mar 31 03:28:01 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (emergency) 
could not position JRC_CSDC_Full.002 to file 147, record 760
Mar 31 03:28:01 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (warning) 
\\.\Tape0 moving: fsr 759: A tape access reached a filemark. 
Mar 31 03:28:01 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (warning) 
\\.\Tape0 moving: fsr 759: A tape access reached a filemark. 
Mar 31 03:28:01 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (notice) volume 
"JRC_CSDC_Full.002" on device "\\.\Tape0": error decoding block, check system 
device configuration, disabling forward space record
Mar 31 03:28:02 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (info) can not 
read record 760 of file 147 on 4mm tape JRC_CSDC_Full.002
Mar 31 03:28:02 jrcn-csdc-is02.ric.pmu.com: NetWorker media: (warning) 
\\.\Tape0 moving: fsr 759: A tape access reached a filemark. 
... this line repeats for 43 times ...
then the next line is when I restarted NSR.

The event log was:

Type: Error

Description:
The driver detected a controller error on \Device\ScsiPort3

Well, ScsiPort3 points to \\.\Tape0 of the Autochanger.

The drwatsn32.log file is dated 03/31/97 3:28:07

I would expect NSR to log a bad tape, but not stop and not create a DR Watson 
file.  I would like NSR to unload the bad tape, flag it a  bad tape, and go 
to the next tape. 

One thing that we had setup incorrectly was that \\.\Tape0 was incorrectly 
set to 4mm and not dlt.  All the other tapes, \\.\Tape1 \\.\Tape2 and 
\\.\Tape3 were set to dlt.  NSR filled up the first tape without any problems 
but complained on the second tape.

Thanks in advance,
Allen Curle

Dr. Watson appearing is one of those feature/bugs
for Windows NT.  I'm not sure we can suppress that
for specific events but willing to send a request
to Legato if you like.

Regarding the potential bad tape: if the tape really
is bad, then we don't have anything to 'solve', per se.
Since we can't fix the tape.  It should not have had
a problem with the 4mm/dlt assignment, that affects
the default capacity used for the tape, not how NSR
treats the tape.  It is important to keep following
good maintenence practices (clean drives regularly,
watch out for pattern of problems, etc).  It isn't
clear to me that the tape is bad, can another drive
use the tape (i.e. could the drive be having troubles)?
(NT doesn't have great diagnostics in this regard as
far as I know).  If there is a sporadic hardware 
problem, it will be harder to determine on NT (vs.
Digital UNIX).

We replaced the tape and everything seems to be working OK now.  I will try 
your suggestion on using another drive for that tape.

Thanks,
Allen