🚀 go-pugleaf

In v8i110, Eduardo Krell says that he has had little or no problem with
APUNIX's "mdump" and other utilities, and that the drive works for him
very well. I'm glad he has had such good luck, but my experience has been
quite different. I won't bore you with a long, tedious story which no one
will read anyway, but suffice it to say that after about 5 months of
hassles, we have only been able to run two full dumps -- and we are having
nothing but trouble trying to reload the files!! (they were dumped using
"mdump".) Our configuration is a 3/160 running 3.5 serving five 3/50
clients with approx. 540Mb of disk space. We have the Exabyte connected to
a Ciprico controller board, and we have even received an extra drive, an
extra Ciprico board, and an extra SCSI cable to go between them so that we
can swap out one or more pieces of the system to see if we have a bad
piece of hardware!! 

So far, no matter what we try, when I try to restore a certain file using
"settape" and "restore ivf /dev/rxb0", the drive runs for a while and then 
just stops. After about 15 minutes, the restore process finally bombs with
the error (don't remember exact words here):"SCSI timeout: error skipping
inode XXXXX", where XXXXX is always a different inode number. I also get
a console message saying: 
"xb0: SCSI select timeout cs=c0<CC,ERR> ss=0x0 (good status) ds=0 (no error)"
This happens with either drive, regardless of the system load. Has anybody
else had this problem? If so, what did you do about it? 

Kerry M. Gough
NASA-Langley Research Center
Mail Stop 149
Hampton VA 23665-5225
(804)864-2496
kerry@teb.larc.nasa.gov
goof@uxv.larc.nasa.gov