Thanks for your good question.
I think it's better to discuss the goals of performance improvement together. So if you have some questions about performance, such as: Which key metric you focus on? How many task throughput you wish to run per minute? Welcome to tell us by reply mail or GitHub issue[1]. [1]: https://github.com/apache/dolphinscheduler/issues/6666; Best Wishes！— ShunFeng Cai ------------------ Original ------------------ From: "dev" <[email protected]>; Date: Fri, Nov 5, 2021 02:05 PM To: "dev"<[email protected]>; Subject: Re: [DSIP][Performance Improvement] Performance improvement for DolphinScheduler 2.0 Well done, Shunfeng! Looking forward our performance improvement. BTW, did we have some goal about it, after this DSIP finish. Best Wish — Jiajie > On Nov 5, 2021, at 12:04, 逆向顺风 <[email protected]> wrote: > > Hi guys, > > As we know,&nbsp;DolphinScheduler 2.0.0-alpha had been released days ago. It 20x faster than 1.3.x,&nbsp;and we will keep the performance optimization for it. > We hope DolphinScheduler is more stable and has higher performance, so it is a long-term challenge that needs to be done consistenly. > > > Actually, I already start it by adding DSIP issue [1], which contains six parts:&nbsp; Master、Worker、Registry、DB、Netty and Performance Test. > And it will be a parent issue to track and summarize the optimization issue for ds 2.0 performance. > This is my initial idea, correct me if I wrong, looking forward your feedback. > > If you have some performance test and suggestion&nbsp;&nbsp;for ds, or just&nbsp;interested in performance optimization,&nbsp;welcome to join us by reply mail&nbsp;or GitHub issue[1]. > > [1]: https://github.com/apache/dolphinscheduler/issues/6666; > > > > > Best Wishes！— ShunFeng > > > &nbsp;
