You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have faced performance issues with Lithium when using it on a linux 12-core server
Even though some benchmarks showed good results (approx 5000 requests per second, 0.3 ms response time), a real conditions scenario resulted in long response times and difficulties with keeping up with the incoming requests
The main difference between the two scenarios is that each request is made by a new connection in the second scenario whereas the benchmarks use the same 200 connections during the whole process
Also, when using 12 threads (by default), the cpu is barely used. Some tests using up to 800 threads improved the benchmarks but did not solve the problem in the real conditions scenario
Description
I have faced performance issues with Lithium when using it on a linux 12-core server
Even though some benchmarks showed good results (approx 5000 requests per second, 0.3 ms response time), a real conditions scenario resulted in long response times and difficulties with keeping up with the incoming requests
The main difference between the two scenarios is that each request is made by a new connection in the second scenario whereas the benchmarks use the same 200 connections during the whole process
Also, when using 12 threads (by default), the cpu is barely used. Some tests using up to 800 threads improved the benchmarks but did not solve the problem in the real conditions scenario
Code
The server used in those tests is from this code
The text was updated successfully, but these errors were encountered: