Experimental Physics and
Industrial Control System

"Kasemir, Kay" <[email protected]> · Tue, 30 Aug 2016 15:39:54 +0000

Hi:

On Aug 30, 2016, at 10:43 AM, Guyotte, Greg S. <[email protected]> wrote:
> I’m experiencing unexpected timeout behavior on PyEpics 3.2.4.
> 
> Using epics.PV, I experience double the expected timeout in all cases.
> Using epics.caget(), the timeout value seems to be ignored and I
> experience 5 seconds of delay when I ask for less than 5 seconds.
epics.caget() is a shortcut which internally calls get_pv to get the PV, then calls get() to read a value.
For the first step, the connection, it always uses a 5 sec timeout. For the get(), it then uses the timeout that you pass in.
So that explains your 5 second delay.

Using the epics.PV in principle allows more control over the various steps: Search for channel, await connection (or register a connection callback), ..

For epics.PV, creating the PV doesn’t really wait at all:
>  print "Test 1: using epics.PV"
>  time1 = time.time()
>  p = epics.PV("idontexist", connection_timeout=waittime)
>  time2 = time.time()
>  print "  Time to create non-existent PV= ",time2-time1
This just sends a name search out and returns in a fraction of a second.

>  val = p.get(timeout=waittime, use_monitor=False)
This get() call will then wait for the connection, since we aren’t already connected.
It uses some internal timeout for that.

I guess you could try this to have more control:

# Send search request
p = epics.PV("idontexist”)
# Wait for the connection
p.connect(timeout=..)

# On success, read:
p.get()

-Kay

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System