You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@seatunnel.apache.org by JUN GAO <ga...@apache.org> on 2023/04/21 06:51:24 UTC

Fwd: 【Discuss】About the flow rate control function

---------- Forwarded message ---------
发件人: 孟非 <me...@gmail.com>
Date: 2023年4月21日周五 10:30
Subject: Fwd: 【Discuss】About the flow rate control function
To: <ga...@apache.org>


请高老师转发给dev@seatunnel.apache.org 邮件。

---------- Forwarded message ---------
发件人: 孟非 <me...@gmail.com>
Date: 2023年4月20日周四 17:38
Subject: 【Discuss】About the flow rate control function
To: <de...@seatunnel.apache.org>


Dear all,
About the flow rate control function, do you think the flow rate control
should be set for the job level? Or do you set flow rate control for each
pipeline under job? For example, if a job has multiple pipelines, each
pipeline has the same flow rate (for example, limiting the average
synchronization of N lines per second) after flow rate control is set for
the job. If the flow rate is at the pipeline level, different flow rates
can be set. If you have other ideas, feel free to contribute to the
discussion!

Which of the following locations would you prefer to have speed limit
controls?
A.
job (Task level. After the job level is set, rate limits are set for all
data pipes. This task may have to deal with 1 or more data pipes)

B.
source (Note: Set only at the source end of the data pipe. After the
setting is successful, set the speed limit for the data pipe where it is
located.)

C.
sink (Note: only set at the target end of the data pipe. After the setting
is successful, the speed limit of the data pipe is set according to this
value)

D.
source and sink (Note: both sides can be set, and the speed limit of the
data pipe where they are located is set according to this. If the rate is
different, the speed limit is set according to the minimum value)

E.
Your other suggestions



Yours sincerely,
Monica


-- 

Best Regards

------------

EricJoy2048
gaojun2048@gmail.com