Back to E155 Home Page
Jan, 23 1997
The E155 online program is designed to monitor the experiment in "real time", it should be able to display enough information to be sure that all the apparatus are working properly. This includes beam monitoring and every pieces of the detectors in the two spectrometers. The target and the Moller detectors are monitored by different programs and are generally used by the respective specialists. The analysis program is able to output raw values coming from the experiment but it can perform all the necessary calculations to provide physical values such as tracks, momenta, energies .
This document has two parts, one for people who want to use the online program as it is without changing anything in the code and from account beamtest on the threeUNIX computers dedicated to the online analysis in the counting house (nevertheless, it is possible to log on on beamtest on these three computers (esaw01, esaw02 and esaw03) remotely. The second part is for those who want to change or write pieces of codes to perform some extra tasks (such as calibration for example). This online manual will explain first how to run the program and use the different tools which are provided for monitoring the E155 experiment.
I assumed for this manual that the program is running on a UNIX computer using a c shell (csh or tcsh). All the program has been developed on an IBM AIX 6000 computer. Assuming that the program already works on another UNIX flavor, this manual should apply the same way. Section Running the analysis program.
First go to the proper directory :
- cd spec2 : to analyze the 2. 75 data
- cd spec5 : to analyze the 5. 5 data
- cd spec10: to analyze the 10. 5 data
Then type : run xldb. It will prompt you to give the name of the executable, type $E155BIN/debug. exe. The debugger window will pop up, press "continue" and let it run. In case of trouble check in the debugger window where it crashed and try to figure out the parameters (such as array subscripts . . . ). If Charlie Young or myself are not available, note the run number, if possible the spill number and the maximum of information. The file name where it crashed, the line, variables . . . Anything which might help us to find the bug. A snapshot of the debugger window can be made with xv. Type in any free window xv (there is a presentation, click on the right button of the mouse to get the main window), type grab and follow the instructi ons. Save the graphic file, that might help us to debug.
A lot of things are then written on the screen. There is the full list of the calibration files read by the program, then the input medium which is open, some new calibration files if necessary (depending of the run number, we shall see in the development section how the calibration files are chosen). All this information will be in the log file too. Now, it is time to look at some results .
A nice way to display some histograms is to use paw.
From the same computer (that can be done remotely) and from the same directory as the analysis, type paw (to learn how to use paw, read the manual (no kidding)). Then inside paw type :
That will map the shared memory which contains the pawc common, now all the histograms which are filled by the analysis program are accessible in "real time", so just plot whatever you want.
There are several hundreds of histograms so it would be more convenient to use some macros (kumacs for PAW) for each subsystem. Each system specialist has developed a bunch of kumacs to monitor the hardware.
For all the kumacs that are following, if you do not know the argument list just type exec xxxxx ? (where xxxxx is the kumac name) to get some explanations.
To know more go to :
Monitoring the subsystems