🚀 go-pugleaf

Frank,

We have been having exactly the same problem. Our local DEC office
has been unable to resolve it. Given the infrequency of the
failure on our cluster (once or twice a month) it is really hard
to track down.  Our configuration is smaller of course:
2 8600s, 1 HSC50, 3 TA78s (1 formatter), 10 RA81s.

My personal belief is that the controller can't recover from
certain types of tape data errors, and just gives up. The TU78s
may be having the same problem with long record lengths that
we (FNAL and CLNS) have been seeing with streaming tape drives.

You didn't mention what revision of HSC code you are running.
Although all of our hardware is up to current rev.,
we have not yet upgraded our HSC50 sofware from 2.50 to 3.00,
since it means rearranging the channel cards and several hours of
downtime for the whole cluster.

We now take data and do our analysis using Massbus TU72s (STC 3650s).
They are much more reliable than TU78s, and have no problems whatsoever
with 28kilobyte records.

I hope that this can be resolved, however.

Selden E. Ball, Jr.
(Wilson Lab's network and system manager)

Cornell University                 NYNEX: +1-607-255-0688
Laboratory of Nuclear Studies     BITNET: SYSTEM@CRNLNS
Wilson Synchrotron Lab              ARPA: SYSTEM%CRNLNS.BITNET@WISCVM.WISC.EDU
Judd Falls & Dryden Road          PHYSnet/HEPnet/SPAN:
Ithaca, NY, USA  14853             LNS61::SYSTEM = 44283::SYSTEM (node 43.251)