[Improvement] Default value optimization for 0.8

[jerqi]

Code of Conduct

• [X] I agree to follow this project's Code of Conduct

Search before asking

• [X] I have searched in the issues and found no similar issues.

What would you like to be improved?

We have developed features to improve the performance of Uniffle. But most of them are turned off by default. Now we will prefer to turn on some config options by default.

How should we improve?

No response

Are you willing to submit PR?

• [X] Yes I am willing to submit a PR!

[jerqi]

Maybe we can make these parameters as default value. @leixm These config options are related to you. Could you raise three pull requests to modify them one by one? (1)

rss.server.single.buffer.flush.enabled true
rss.server.single.buffer.flush.threshold 128M

(2)

rss.coordinator.select.partition.strategy CONTINUOUS

(3)

rss.server.max.concurrency.of.per-partition.write 30

[leixm]

Maybe we can make these parameters as default value. @leixm These config options are related to you. Could you raise three pull requests to modify them one by one? (1)

rss.server.single.buffer.flush.enabled true
rss.server.single.buffer.flush.threshold 128M

(2)

rss.coordinator.select.partition.strategy CONTINUOUS

(3)

rss.server.max.concurrency.of.per-partition.write 30

Sure, i will raise PRs later.

[zuston]

I will make the client compress type as ZSTD by default. cc @jerqi

[jerqi]

I will make the client compress type as ZSTD by default. cc @jerqi

It's ok for me if we have verified in the production environment.

[zuston]

I will make the client compress type as ZSTD by default. cc @jerqi

It's ok for me if we have verified in the production environment.

ZSTD has been applied in our production for 1 year by default

[jerqi]

I will make the client compress type as ZSTD by default. cc @jerqi

It's ok for me if we have verified in the production environment.

ZSTD has been applied in our production for 1 year by default

OK.