I would like to nominate fast_pool
a fast async pool based on channel。
You can think of it as a substitute for connection pools such as deadpool and mobc
- support
get()
,get_timeout()
,state()
methods - support atomic max_open(Resize freely)
- Fewer code, clearer logic, less prone to bugs
- based on flume channel
way fast_pool?
- fast get() method performance
//windows:
//---- bench_pool stdout ----
//use Time: 4.0313ms ,each:40 ns/op
//use QPS: 24749412 QPS/s
//macos:
//---- bench_pool stdout ----
// use Time: 6.373708ms ,each:63 ns/op
// use QPS: 15683710 QPS/s