We are currently seeing a problem on our secondary fault tolerant SpectroSERVER. The disk drive is filling up due to the size of the $SPECROOT/SS/.moot.trace file. What can we do to reduce the amount of data logged to the .moot.trace file?
The issue is caused by low thread resources on the SpectroSERVER, from the $SPECROOT/SS/VNM.OUT file, which is triggering the extra data to be logged in the .moot.trace file.
"WARNING at CsWorkSched.cc(173): Low thread resources detected. Work scheduler requesting 1min/50max threads, but not available."
You can also check the VNM model's "SpectroSERVER Control" subview -> "Thread Information" subview to see if the "Total Work Threads" peak out at 4,000 (default value).
This is resolved by the Spectrum_10.01.01.D177 patch, which limits the amount of data logged due to low thread resources. The patch limits the dumps to 10 a day, at 10 MB per dump. Therefore the .moot.trace file should not grow more than 100 MB a day if we detect a low thread condition.
If you suspect you are having this issue, please contact support for a copy of the patch.
This will be addressed in Spectrum 10.3.