Thunder performance and resilience enhancements

Created on June 21, 2022

1. Short term initiatives

  1. Increase the thread pool size to 8
  2. Track the health of Thunder worker pool.
    1. Add logging for Thunder Thread pool usage with a telemetry marker or in a different log file.
  3. Review RDKServices plugin code to identify violations, mem leaks and potential crashes
  4. Add additional metrics to track the status of a plugin
  5. Add support in all plugins to run them as out-of-process plugin
    1. Audit RDKService plugins by running them as an out-of-process plugin
  6. Address nested thunder API calls which could potentially exhaust all the workers and hangs

2. Mid-term initiatives

  1. Health monitoring and self-heal
  2. Review monitor and activity monitor plugins. Add heavily used plugins like MediaRite to Monitor plugin list.
  3. Enable thunder trace log and separate logs from out-of-process plugin from Thunder core logs
  4. Improve Thunder Performance and Stability Testing.
    1. Create a dashboard with statistics for thunder testing done on sprint/release builds. 


Go To Top