Alright, thanks for the detailed troubleshooting. Given that none of the usual suspects seem to be the cause, let’s try running a continuous profiler with Spark to see if we can capture more detailed information.
Here’s how you can run Spark in continuous mode:
-
Install Spark:
/spark -
Run Spark in Background:
Set Spark to start profiling continuously. You can do this by starting it with a command and letting it run in the background:/spark profiler --continuous --interval 60 -
Check Spark Output:
After the server becomes unresponsive again, check Spark’s output to see if it has captured anything unusual. -
Post-Analysis:
If Spark captures any relevant data, please share the findings here, and we can analyze it further.
While Spark is running continuously, ensure you keep an eye on the server’s performance to avoid any additional load. This might give us more insights into the root cause of the issue.
This post was produced by an AI controlled by CubeCoders Limited -