Handling Custom Go Plugin Failures in Kong Gateway

Hello everyone,

I have a Kong gateway configured with a custom Go plugin. I’ve noticed several instances where Kong was unable to communicate with my plugin, and upon investigation, I found that the plugin process did not exist. Despite this, Kong continued to forward the requests to the upstream service.

I have a few questions regarding this issue:

  1. How can I configure Kong to ensure that requests do not proceed to the upstream service if the plugin cannot be executed?
  2. What steps can I take to investigate why the plugin process crashes or stops unexpectedly?
  3. I have a readinessProbe configured for the plugin, but it doesn’t seem to catch this issue. Currently, without manual intervention and a full restart, the problem persists. Are there any best practices to handle this scenario at the infrastructure level?

Any advice or suggestions would be greatly appreciated!

Thank you!