I am running a Python IOx agent that can suddenly hang, mid-execution. In IOx, the application is still shown as running but no more logs are output (or any other form of interact with other services, e.g. http calls to our server).
The App has 2 threads and both are hung.
I can, however, then "bring it back to life" by merely executing an ioxclient app console command for the App. Logging resumes, from some utterly random point in the code (it could have been midway through execution of a function); expected log messages are output with the now current timestamps and all is good once again.
If I exit out of the console session, however, the App will again freeze up at some point in the future.
It seems the frequency of failure can be increased by increasing App logging levels and redirecting output to console.
Solved! Go to Solution.
Thanks for that, Steve. It was indeed the problem - a 3rd party module was slyly logging to console when debug logging was enabled. Overrriding logger settings for that module solved the problem.