Benchmark oatpp - AWS cloud

AWS t2.micro instance. 1vCPU 1GB Ubuntu 18.04 - free tier

Performance test of "Hello World" endpoint.

Machine - AWS t2.micro instance. 1vCPU 1GB Ubuntu 18.04 - free tier
Tool - wrk

oatpp-async compiled with: -D OATPP_ASYNC_HTTP_CONNECTION_HANDLER_THREAD_NUM_DEFAULT=1 option.

benchmark aws t2.micro

  • go net/http service performs clearly till 10 concurrent connection. At higher concurrency levels performs with "timeout" errors. Down at concurrency level > 15K.
  • oatpp-async service performs clearly and stays available up till 20K concurrent connections load. At higher loads no response. Service available if load drops.
  • oatpp-multithreaded service shows very high performance and works clearly up to 2.5K concurrent connections. Then down at load 5K concurrent connections.

AWS t3.xlarge instance. 4vCPUs 16GB Ubuntu 18.04 - $121.81/Month

Performance test of "Hello World" endpoint.

Machine - AWS t3.xlarge instance. 4vCPUs 16GB Ubuntu 18.04 - $121.81/Month
Tool - wrk

oatpp-async compiled with: -D OATPP_ASYNC_HTTP_CONNECTION_HANDLER_THREAD_NUM_DEFAULT=3 option.

benchmark aws t3.xlarge

  • go net/http service performs clearly till 10 concurrent connection. At higher concurrency levels (till 25K connections) performs with "timeout" errors. At 30K load no response. Service available if load drops.
  • oatpp-async service performs clearly and stays available up till 25K concurrent connections load. At 30K load no response. Service available if load drops.
  • oatpp-multithreaded service shows very high performance and works clearly up to 2.5K concurrent connections. At higher concurrency levels (till 20K connections) performs with "timeout" errors. Down at higher load.