Experimental Physics and
Industrial Control System

Dave Reid <[email protected]> · Thu, 17 Dec 2009 10:33:24 -0800

Ralph Lange wrote:

any particular reason for running each soft IOC on a box of its own? I

know people are happily running a dozen or more on a pretty standard box

- you could declare most of your inventory spare.

Not objecting, just curious...

Hey Ralph....

We had to sit here for a few minutes to try and remember why we didn't go that

direction.

The reason is that in our initial testing, we found problems with the timeliness

of sending out Channel Access messages from the IOC if there were multiple IOC's

running on the same CPU.

More specifically, we were controlling a power supply through a knob interface

that we developed, and if there were other IOC's running, such as one

communicating periodically with our modicon PLC, control messages to the power

supply would "stack up".

Also, we decided that we wanted to compartmentalize our IOC's by fuctionality.

That way, if we lose a CPU for some reason, we only have one IOC that that stops

working.

dave

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System