Experimental Physics and
Industrial Control System

Ric Claus <[email protected]> · Fri, 21 Feb 1997 14:57:12 -0800

Sorry for the delay in getting more information out to tech-talk about
our 
hanging IOC problem.  I was waiting for more statistics.  Thanks to 
everyone for supplying suggestions on what to look for.  At this point
I'm fairly confident (but more statistics would help) that the problem
is
not due to memory exhaustion or too many CA clients.  I think instead
that
tNetTask is dying on a flood of messages being sent to our message
logger.
That, in turn causes everything else to lock up to the point where there
is
no response from the tShell task.  Even ^X is ignored and the only way
out
is to push the reset button.

Jeff Hill postulates that perhaps the NI VXIcpu030 hardware is a
problem.
Jeff had alerted me to a rev level problem several years ago and so I
have
ensured that all our NI VXIcpu030 boards are rev D7 or higher.  The
latest
round (and the unit in question) is rev level E1, if I'm decoding the
board's
marking correctly.  At this point, I think that the cpu030 is a pretty
robust
board.

On a completely different subject, today is my last day at SLAC
(although
I'll be consulting for a while to come yet).  I want to say that I've
really 
enjoyed the contact that I've had with many of you in the EPICS
community.  
Thanks for a neat product and the opportunity to learn about its details 
from its authors.  I wish you lots of luck and success with it in the
future.

	Cheers,
		Ric
-- 

	Ric Claus	([email protected], (415) 926-2697)

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System