Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-765199553
Waiting for the test result!
This is an automated message from the Apache Git Service.
To respond to the message,
Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-764518459
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use
Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-765098590
Cool! I think we need some data to prove the point
This is an automated message from the Apache Git Service.
To
Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-764605634
@idbeta How could we know that "not differ too much" would happen, or "they
are within an order of magnitude". Maybe we need some methods to get over those
false positives, but not
Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-764530183
> ok, good idea, the performance of Github action will change a lot, I
tested it several times, QPS from `13300` to `21433`, so different.
Wow that indeed differs a lot. Would
Yiyiyimu commented on issue #3391:
URL: https://github.com/apache/apisix/issues/3391#issuecomment-764518459
The lowest result out of 5 test seems very easy to be surpassed. Maybe we
could test for like 100 times, get its distribution and try to set the
benchmark value