Performance Degradation in Enterprise Gateway with Multiple Kernels

[zdhdsx]

Hi everyone,
I'm facing a significant performance issue with my Enterprise Gateway setup when multiple kernels are launched. Specifically, the response time for a simple health check endpoint has jumped from 1ms to 10 seconds. Restarting the gateway temporarily fixes the problem, but the performance degradation returns as more kernels are started.
I've checked the CPU and memory usage, and both seem to be within normal ranges. However, I’ve noticed that the number of TCP connections keeps growing over time, which might be a clue to what’s causing this.
Here’s some information about my setup:
The Enterprise Gateway is running on a single node.

The kernels are also running locally on the same machine.

The performance drop becomes noticeable after starting around 20 kernels

Has anyone run into a similar issue before? Could you suggest possible causes or solutions? I’d also appreciate any tips on how to further troubleshoot this problem.
Thanks so much for your help!

[zdhdsx]

kernel is working on k8s