We have the "Contact Lost Model Destruction" in Spectrum enabled and configured to delete models that have been in a lost contact status for more than 30 days.
We are seeing models in Spectrum that have had lost contact status longer than 30 days not being deleted as expected.
Release : Any
For this specific issue, it was found there is a serious performance issue in the SpectroSERVER causing significant latencies in the work threads.
If you are experiencing performance issues in Spectrum, please open a case with Spectrum Support to investigate and resolve the root cause of the issue.
To get a head start when opening a case for a performance issue in Spectrum, please collect perfCollector9 and moot trace data and upload to the case.
To run perfCollector9:
1. Log into the SpectroSERVER system as the user that owns the Spectrum installation
2. If running Windows, start a bash shell by running "bash -login"
3. cd to the $SPECROOT directory
4. Remove the Performance directory if present
5. Execute the following command where <HOST> is the hostname of this system:
./perfCollector9 <HOST>
6. When complete, upload the $SPECTRUM/Performance/results.tar.gz file to the case for analysis
To run a moot trace:
1. Log into the SpectroSERVER system as the user that owns the Spectrum installation
2. If running Windows, start a bash shell by running "bash -login"
3. Find and note the process id of the SpectroSERVER process.
4. Run the moot trace where <PID> is the process id of the SpectroSERVER process noted in the previous step:
a. In Linux, run the following command:
kill -USR1 <PID>
b. On Windows, cd to the $SPECROOT/lib/SDPM directory and run the following command:
./kill -USR1 <PID>
5. Leave the moot trace running for 15-20 minutes
6. Rerun the command to stop the moot trace
7. Upload the following files to the case:
$SPECROOT/SS/.moot.trace
$SPECROOT/SS/VNM.OUT
All of the $SPECROOT/SS/support/Incident.Data files