This is a clone of issue OCPBUGS-216. The following is the description of the original issue:
—
Description of problem:

In customers env we've seen kuryr-controller being restarted due to timeouts of the liveness probes. We default that to 1 second, which for bigger and busy envs might be too little.

Version-Release number of selected component (if applicable): 4.8

How reproducible: On bigger envs.

Steps to Reproduce:
1. We've seen that with 1500 pods and 6000 Neutron ports.

Actual results: kuryr-controller is restarted, `context-deadline-exceeded` in the pod events.

Expected results: kuryr-controller keeps running.