Dear Everyone,
I have been tasked with fixing an error in a software product that my company sells and I desperately need some help. The software is used for controlling a machine and uses serial (rs-232) communication for doing so. The PC and the device continuously exchanges data and if either end detects that no new data has been received with in a 5 second timeout period the connection is dropped. Some users complains that the connection is dropping rather frequently (2-3 times per 7-8 hours). What happens is that the serial port stops working for a few seconds (enough to trigger a disconnect). This happens with both built-in serial ports and USB-to-serial adaptors.
I have been doing a lot of testing to see if I could reproduce the problem or come up with a generally applicable solution. I have tested with 4 different Windows 7 PC's and three different serial ports. On three of the four machines I have been unable to reproduce the error but on the last one I found an interesting thing. When I start a stress test program (BurnInTest) the machine freezes for about 5 seconds. This does not happen on my other test machines. It turns out that this program uses Windows Management Instrumentation for detecting the hardware setup on the machine and I suspect that this somehow causes the freeze. The event viewer does not report any errors in the WMI log.
I theorize that this problem or a similar one could cause the disconnects. Most of the users of our product are only using the PC sporadicly and will probably not notice if the UI freezes for a few seconds once or twice during the day and I have yet to witness a breakdown myself because of the randomness. This makes diagnosing the problem that much harder. I have a few questions that I would like some assistance with:
Is there a tool for detecting temporary lock-up's on the system level and on the driver level?
What might cause such temporary lock-ups and how should I proceed to diagnose the underlying problem?
Thank you
I have been tasked with fixing an error in a software product that my company sells and I desperately need some help. The software is used for controlling a machine and uses serial (rs-232) communication for doing so. The PC and the device continuously exchanges data and if either end detects that no new data has been received with in a 5 second timeout period the connection is dropped. Some users complains that the connection is dropping rather frequently (2-3 times per 7-8 hours). What happens is that the serial port stops working for a few seconds (enough to trigger a disconnect). This happens with both built-in serial ports and USB-to-serial adaptors.
I have been doing a lot of testing to see if I could reproduce the problem or come up with a generally applicable solution. I have tested with 4 different Windows 7 PC's and three different serial ports. On three of the four machines I have been unable to reproduce the error but on the last one I found an interesting thing. When I start a stress test program (BurnInTest) the machine freezes for about 5 seconds. This does not happen on my other test machines. It turns out that this program uses Windows Management Instrumentation for detecting the hardware setup on the machine and I suspect that this somehow causes the freeze. The event viewer does not report any errors in the WMI log.
I theorize that this problem or a similar one could cause the disconnects. Most of the users of our product are only using the PC sporadicly and will probably not notice if the UI freezes for a few seconds once or twice during the day and I have yet to witness a breakdown myself because of the randomness. This makes diagnosing the problem that much harder. I have a few questions that I would like some assistance with:
Is there a tool for detecting temporary lock-up's on the system level and on the driver level?
What might cause such temporary lock-ups and how should I proceed to diagnose the underlying problem?
Thank you
