Difference between revisions of "Epics Logger Scripts for SBS"
From Hall A Wiki
| Line 2: | Line 2: | ||
the logging of information for the SBS. | the logging of information for the SBS. | ||
| − | == [[Explaining the EPICS logging scheme for SBS | + | == How to Add Variables to EPICS logging for SBS == |
| + | |||
| + | <pre> | ||
| + | 1. Edit EpicsToFile, add your variable. | ||
| + | 2. Might want run this with TESTPRINT=1 and run EpicsToFile interactively to check it. Don't forget to restore TESTPRINT=0. | ||
| + | 3. Kill the current running of EpicsToFile | ||
| + | |||
| + | [sbs-onl@adaq1 halla]$ ps awx | grep EpicsToFile | ||
| + | 4018 pts/69 S+ 0:00 grep --color=auto EpicsToFile | ||
| + | 10161 ? S 1:03 /bin/ksh /adaqfs/home/sbs-onl/logentry_scripts/halla//EpicsToFile | ||
| + | |||
| + | kill -9 10161 | ||
| + | |||
| + | 4. Then wait. The cron script "startCron" will restart it in a few minutes. Perhaps check the process list to be sure. | ||
| + | </pre> | ||
| + | One fly in the ointment at the moment is that if someone is doing "less EpicsToFile" or "more EpicsToFile" | ||
| + | on adaq1, then unfortunately startCron is not smart enough to restart EpicsToFile. It is smart enough | ||
| + | to ignore emacs sessions. In the meantime, kill the "less" command, too. | ||
| + | |||
| + | <pre> | ||
| + | [sbs-onl@adaq1 halla]$ ps awx | grep EpicsToFile | ||
| + | 2312 pts/69 T 0:00 less EpicsToFile | ||
| + | 3034 pts/69 S+ 0:00 grep --color=auto EpicsToFile | ||
| + | 10161 ? S 1:04 /bin/ksh /adaqfs/home/sbs-onl/logentry_scripts/halla//EpicsToFile | ||
| + | |||
| + | kill -9 2312 | ||
| + | kill -9 10161 | ||
| + | |||
| + | </pre> | ||
| + | |||
| + | |||
| + | == Explaining the EPICS logging scheme for SBS == | ||
| + | |||
The EPICS data go to two places | The EPICS data go to two places | ||
| Line 9: | Line 41: | ||
2. datastream as event type 131 | 2. datastream as event type 131 | ||
| − | + | Let me describe the scheme. | |
A script called EpicsToFile runs continuously, using caget to obtain EPICS | A script called EpicsToFile runs continuously, using caget to obtain EPICS | ||
| Line 42: | Line 74: | ||
and getruninfo_bgr_SBS is called by | and getruninfo_bgr_SBS is called by | ||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
| − | |||
Revision as of 17:39, 23 October 2022
Here is some technical information about how to add EPICS variables to the logging of information for the SBS.
How to Add Variables to EPICS logging for SBS
1. Edit EpicsToFile, add your variable. 2. Might want run this with TESTPRINT=1 and run EpicsToFile interactively to check it. Don't forget to restore TESTPRINT=0. 3. Kill the current running of EpicsToFile [sbs-onl@adaq1 halla]$ ps awx | grep EpicsToFile 4018 pts/69 S+ 0:00 grep --color=auto EpicsToFile 10161 ? S 1:03 /bin/ksh /adaqfs/home/sbs-onl/logentry_scripts/halla//EpicsToFile kill -9 10161 4. Then wait. The cron script "startCron" will restart it in a few minutes. Perhaps check the process list to be sure.
One fly in the ointment at the moment is that if someone is doing "less EpicsToFile" or "more EpicsToFile" on adaq1, then unfortunately startCron is not smart enough to restart EpicsToFile. It is smart enough to ignore emacs sessions. In the meantime, kill the "less" command, too.
[sbs-onl@adaq1 halla]$ ps awx | grep EpicsToFile 2312 pts/69 T 0:00 less EpicsToFile 3034 pts/69 S+ 0:00 grep --color=auto EpicsToFile 10161 ? S 1:04 /bin/ksh /adaqfs/home/sbs-onl/logentry_scripts/halla//EpicsToFile kill -9 2312 kill -9 10161
Explaining the EPICS logging scheme for SBS
The EPICS data go to two places
1. HALOG -- both start-of-run and end-of-run.
2. datastream as event type 131
Let me describe the scheme.
A script called EpicsToFile runs continuously, using caget to obtain EPICS variables and put them into text files.
File to log EPICS data /adaqfs/home/sbs-onl/logentry_scripts/halla/EpicsToFile
the text files are in /adaqfs/home/sbs-onl/logentry_scripts/halla/epics/halla.epics
Note, halla1.epics is rewritten frequently while halla.epics is a more stable copy that gets put into HALOG and datastream.
For the HV, its similar. A script /adaqfs/home/sbs-onl/logentry_scripts/HV_logging_SBS/EpicsLooper with data in the /Results subdirectory.
Other scripts insert the halla1.epics file into the datastream and into the HALOG.
All the scripts mentioned are in the directory /adaqfs/home/sbs-onl/logentry_scripts/halla/
InsertETLoop -- loops forever, puts halla.epics into ET system as event type 131 getruninfo_bgr_SBS -- inserts EPICS data into HALOG. and getruninfo_bgr_SBS is called by
