3 ms·
No, you still have a CPU request, and the HPA is based on utilization of that. Also, even without limits I believe CPU is prioritized based on the request. So
by sciurus 6y ago
No, you still have a CPU request, and the HPA is based on utilization of that.
Also, even without limits I believe CPU is prioritized based on the request. So if 1 pod requests 100 millicpu and another pod requests 200 millicpu, if they both try to use all the CPU on a node the one that requested 200 millicpu will use 2/3 of the CPU and the other will use 1/3.