Experimental Physics and
Industrial Control System

"Lawrence T. Hoff" <[email protected]> · Thu, 01 Mar 2007 16:35:25 -0500


	At RHIC, we actually developed a server to
perform that function. The server establishes connections
to all the terminal servers (digis), and logs all output.
	It also periodically adds a timestamp to the
log file (between lines of output) to help correlate
messages with "external" events. It also does various
bits of file management (different log file for each
IOC, new log file each 24 hours, date included in
log file name to help searching, etc.)
	It accepts user "logins" (up to 10 per IOC)
and directs user I/O to/from the terminal server.
I.e. it allows for user interaction with the IOC console
port concurrent with data logging, and/or multiple "live"
screens looking at IOC output concurrent with logging.
	Finally, we connected the server to the DB
which indicates operational status of the IOC, so
we can selectively enable/disable IOC participation
by changing the DB entry (e.g. disable during periods
of active development).
	That last feature makes this server fairly
facility-specific, so I don't think it is helpful to
offer to share the code. Different sites probably
have different ideas about how to perform the file
management and/or timestamping anyway. If any of these
ideas sound promising for you, you could probably code
up something similar in short order.
HTH -- Larry
P.S. Sometimes systems go down even without a "last gasp".
E.g. we used to have VME "power fail modules" which failed
by continually asserting SYSRST. This does not produce any
"last gasp", and could only be remedied with a power cycle,
and only cured by replacing the "power fail module".

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System