attempt to put scripts in CODA, failed, backed out

NEXT
Make New Entry, Make Followup Entry

User name R. Michaels

Log entry time 09:14:21 on February10,2011

Entry number 346254

keyword=attempt to put scripts in CODA, failed, backed out

This morning at 08:00 I stopped the cosmics run (looks like you
had plenty of data, about 7 filesplits) and tried to put in the
scripts we need for CODA. This failed because the scripts apparently
hang up state transitions even if they are put in the background.
It worked for CODA 2.5 (previous version) but so far does not work
for this newer version 2.6.1. So, I've backed out of this, i.e.
removed the scripts. So the DAQ is working again.

I've sent an e-mail to coda@jlab with explicit details of the
symptoms, so hopefully we can resolve this. Running the scripts
is considered essential. I guess one can run them seperately
from CODA, but we cannot rely on shift workers to remember to do
that. Hopefully the DAQ group has a solution. FWIW, here's
my e-mail to coda@jlab. If anyone has an idea let me know.

-----------------------------------------------------------

We notice that for a script tied to EB, it hangs up
CODA forever, even if the script puts all it's work
into the background.

For example, I call "getruninfo", which in turns does this

$SCRIPT_DIR/getruninfo_bgr &
exit 0

with CODA 2.5 this worked fine -- the script getruninfo_bgr
went into the background, and CODA ran fine. But with
CODA 2.6.1 it hangs up with a message repeated infinitely
in rcgui:

Script (....)
Waiting for EB2 (repeated infinitely)

This script needs to be put in the background because, for
example, it logs EPICS and scaler info throughout the run.
So it needs to run concurrently with data collection.

I tried tying the script to ROC1 instead. Same thing
happens:

Waiting for ROC1