How to correctly configure the concurrency of a single function?

We hardware is 12vCPU and 32GB RAM K8s cluster, 3 nodes.
My problem is that whenever I try to configure the container concurrency, such as setting the `-c` parameter to 2, 5, 10, 50, etc., but now it seems that OpenWhisk can only work correctly when the concurrency is 1.
The specific performance is that the error of preheating the container directly occurs in the function call, as follows：
And the test program also shows a large number of timeouts, and this test runs normally when `-c` is configured as 1, I don’t know how to set up concurrency：
I have configured value.yaml according to the suggestion you gave in the previous issue, as follows：
At the same time, I also found an obvious problem, that is, when the number of preheated containers reaches 200, it is difficult to configure upwards, and there will be an error that the api gateway resource cannot be found. This is in the previous issue.

My current difficulty is that it is difficult to use OpenWhisk to cope with a large amount of load. The data I currently get is only 200TPS, which is a very bad value. My function is very simple, just remotely calling a python3 runtime function of a redis database.
Could you give me some advice?

No source files or tests are named. Start with the `-c` concurrency setting and the configured `value.yaml`, then reproduce the Redis-backed Python function load while examining the preheated-container and API gateway resource errors; done means concurrency above 1 can sustain the reported load without timeouts or preheating failures.

Written by the indexing model from the issue text.