Allows the ability to turn the read wait time off or tweak the settings for your own system.
I benchmarked this against a local EPP server. Added a 1 second delay to each request (server side) to simulate longer response times to trigger the cpu usage issue.
For reviewer: Please review, run tests and check if this works properly. I have benchmarked and tested this manually but perhaps your unit tests might reveal some issues.
Allows the ability to turn the read wait time off or tweak the settings for your own system.
I benchmarked this against a local EPP server. Added a 1 second delay to each request (server side) to simulate longer response times to trigger the cpu usage issue.
For reviewer: Please review, run tests and check if this works properly. I have benchmarked and tested this manually but perhaps your unit tests might reveal some issues.
The benchmarking data:
Lower loops = lower cpu usage = better
no usleep
current (master) sleep at 100 microseconds
current but longer sleep (1ms)
this new branch
This new branch but set Incrementor limit to 50000