-
Notifications
You must be signed in to change notification settings - Fork 26.4k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Triple协议性能优化 #10558
Comments
dubbo版本3.1.0 jdk8和jdk17性能差距也非常大
grpc的unary单线程阻塞模式下: 8K ops/s
grpc的unary单线程阻塞模式下: <7.1K ops/s |
当前进度:
|
当前进展:
|
基于 @icodening 的思路减少多次flush,合并flush的模式
tri的unary单线程阻塞模式下:5.7K ops/s |
已经的准备提交的features
今天做了几个测试
|
问题记录 2023.2.27经排查,发现Triple性能与GRPC相差还是很大的原因为Netty自身过多的flush导致. 本地修复后加上#10754 的改动,benchmark结果大幅提高已接近GRPC. 结果如下 triple pb优化前(3.2.0-beta.5-SNAPSHOT)
triple pb优化后
grpc
netty本地改动点由于涉及netty源码改动,triple侧性能调优待定。 |
Dubbo Beanchmark 记录测试时间: triple pb 3.2 优化后
dubbo grpc 3.1.7
triple pb 3.1.7
triple pb 3.2 优化后
triple wrapper 3.2
dubbo fastjson2 3.2
dubbo hessianlite 3.2
dubbo pb 3.2
|
@icodening 每次跑完测试,除了原始数据外,最好能加一个结论:相比于 grpc, dubbo/triple 的性能提升/下降 x% ,如果能再有几张对比图就更好啦 |
好的,我看剩下的PR也已经合并了,我找个时间跑一下合并后的版本再梳理一下数据吧 |
The text was updated successfully, but these errors were encountered: