Hi everyone,
I’ve been working with Kong Gateway in a local development environment for some performance testing, and I’ve run into a strange issue regarding request latency and plug-in stability that I’m hoping the community can help me clarify.
Recently, I’ve noticed that whenever my host system is under heavy resource load—specifically when I’m testing how an arceus x executor handles high-concurrency background scripts—the response times through the Kong proxy start to spike significantly. It’s not just a slow connection; I’m seeing the rate-limiting plugin trigger prematurely on certain routes, even though the request volume hasn’t actually increased. It feels like the gateway is struggling to accurately track the request time-windows when the CPU is saturated by these other background processes.
While discussing this with a few colleagues, we wondered if Kong’s internal Nginx workers might be experiencing context-switching delays or if the shared memory zones used by plugins are becoming unresponsive when the system is pushed to its limits. Has anyone else noticed their local gateway instances hitting silent timeouts or experiencing “jagged” performance metrics when their machine is working hard on intensive background automation?
I’m trying to figure out if there’s a way to ensure Kong’s worker processes maintain higher priority, or if I should be looking into adjusting the mem_cache_size and worker_processes settings to be more resilient during these high-multitasking scenarios. If you’ve found a way to keep your API traffic fluid even when the host system is under heavy background load, I’d really appreciate your insights!
Best regards!