Mobile cFP-2020 hangs

[Sorbens]

Hi!
Background: Mobile Datalogger for a transportation cooling system including temperaturs, humidity, fuel, GPS / GSM Modem (serial ports) with a cFP-2020/LabView system.
I have different loops for reading sensor values, write data to flash, communicate with GPS and Modem via RS232.
Problem: CPU hangs after a couple of hours every time it is rebooted.

Does anybody have an idea?

Regards, Louis

[Matthew_Williams]

First guess is a memory issue. Make sure you are pre-allocating all of your arrays, replace array
elements instead of inserting, etc. Have you monitored memory usage over the time that it does run?
Does is always crash at approximately the same run time?

Matt

[Aaron_G]

Louis,

What do you mean by hang? Does the program simply appear to stop doing anything, or does the unit re-boot and start flashing the status LED?

Assuming that it is the latter (which is a crash), the most likely culprit is memory allocation. Assuming that it is the former, than you will have to add in diagnostics to assist you in tracing the source of the hang. A question that comes to mind is are you using the Error cluster, and if so, how are you handling errors? If you wire the error cluster to all of the VIs, but do nothing to handle it, then after an error occurs, most VIs will stop executing (if Error Status==True).

I recently had a program that appeared to hang after 24-48 hours. The system had 8 parallel loops each running a state machine and only some were getting hung. I added debug information to each loop through by using the PublishData.vi and published separate information on the state of each loop. By watching the timestamps and information, I was able to determine in which states the loops were when the hang occurred, and then added more debug markers in the specific states to further isolate the cause. I repeated the process until I had identified the specific VI that caused by hang. You can also use the LEDs (A-D) to display status information of the program.

Regards,
Aaron

[Sorbens]

Thanks for your help!

I was carefully with handling errors and I can't imagine a memory problem - but I will check all arrays in the VI's.
Non of the loops is working, the CPU is not doing anything anymore. I don't know yet, if there is a certain crash time, because the Logger is far away and I have only the information from the Flash-Card (once a week)/ Modem.
The strange to it is that I have a copy of the system (cFP-2020, temperature, counter, analog module) in the laboratory with the difference that there is no GPS, no Modem - beside this the same software, same amount of measurement channels. It works without any problem!
Therefor I could imagine some hardwaredefekt or a problem with the operating system handling the serial ports.
Louis