Re: Exploring EPICS performance/processing limits

[Ralph Lange via Tech-talk <tech-talk at aps.anl.gov>]

Hi Marco,

Please note also that what you're showing is testing a whole chain: python - CA client - network - CA server - IOC database - stream device - asyn - loopback driver.

Lots of parts and lots of failure modes.

If you want to check only the PV processing time (i.e., from the database down), I would add a database on the same IOC that sets the values. Think calc records as counters, seq records for small sequences of things, etc. That would leave out the Channel Access and python client layers and maybe allow locating issues easier.

Cheers,
~Ralph