20 Nov
2014
20 Nov
'14
12:22 p.m.
I have to say this about the LCM VAX: it successfully ran CMUIP for a
rather long time. The pattern of failures, becoming more and more
frequent, seems to conform to a hardware issue rather than one of
software. The machine has never had a large load - when I watched it
regularly, there were rarely more than a handful of users at any one time.
The machine has been power-cycled in response to errors, which would of
course reinitialize all transient data structures - and I do not believe
that CMUIP uses persistent caches (i.e. cached to disk).
I have this amazing feeling of deja-vu all over again :-)
Here, CMU/IP ran ok for ages but then died every few hours for a couple of
days. Then it ran ok for another few weeks or months.
We blamed the hardware and gave the resident DEC engineer a pain over it.
We ran CMU/IP on several machines and that it seemd to die on some but not
on others lent weight to the hardware theory but there was nothing wrong with
our hardware. (We also blamed network problems, and there were network
problems but they were nothing to do with CMU/IP falling over.)
Our former colleague who had jumped ship and went to work for an outfit with
money had ditched CMU/IP and pleaded with us to see the light and do the same
but we persisted in trying to debug CMU/IP or to place the problem elsewhere.
He had the same problems convincing us that I am having now.
The machines where it wasn't falling over had lousy performance and were
painful to use interactively with delayed and jerky echoing and general grief,
whether anyone was logged in via CMU/IP or not. All this went away when
CMU/IP was removed. It was a revelation to us how smooth and responsive the
machines could be.
I guess the museum is doing it's job perfectly - it's giving me a vivid
recreation of something I went through 20 years ago :-)
I hate to pour scorn on free software that people have obviously put a huge
amount of effort into producing but it was not ready for prime time 20 years
ago and it has not changed since. One of the big strengths VMS has is
reliability. CMU/IP puts a serious dent in that. It also defeats the logical
approach to problem solving that VMS normally promotes and remains one of the
very few VMS applications that requires reboots to untangle it. If ever there
was a case for for putting a sofware exhibit in a glass case and looking at
it rather than using it, CMU/IP is it.
Yes, LCM could just load UCX and perhaps whistle a happy tune. It might be
an interesting experiment to do so and observe behavior - the changes in
the startup script could easily be commented in/out. I'd certainly like to
see them continue to use CMUIP for historical reasons. Multinet would also
be interesting, if policies have changed at LCM to provide for licensing
costs of that software. When I was restoring the machine back in 2008-9,
Process Software offered a somewhat amusing 'discount' for an educational
institution. -- Ian
Check the error log first, just in case.
I don't know whether UCX is compatible with happy tunes. Early versions did
not enjoy a good reputation but maybe it has improved. Still, it's got to be
better than CMU/IP and it might exonerate the hardware.
It might be worth contacting Process again. Things might have changed over
the last five years. Maybe they might be open to a special deal for a
contemporary version of Multinet on a museum exhibit if your machine is not
also used for more than that?
Regard,
Peter Coghlan.