GitHub user 10110346 opened a pull request: https://github.com/apache/spark/pull/22065
[SPARK-23992][CORE] ShuffleDependency does not need to be deserialized every time In the same stage, 'ShuffleDependency' is not necessary to be deserialized each time. I have tested 3 times in my production environment , it has a bit of performance improvement( about 0.7%): Before this PRï¼ duration: 20189(s) App count: 4736 the sum of time for all Apps: 530470(s) Afterthis PRï¼ duration: 20035(s) the count of Apps : 4736 the sum of time for all Apps: 525220(s) How was this patch tested? Existing unit test You can merge this pull request into a Git repository by running: $ git pull https://github.com/10110346/spark notdeserializedep22 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/spark/pull/22065.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #22065 ---- commit a99769dd1aac779e972ed2e23aa7598e6d7c7105 Author: liuxian <liu.xian3@...> Date: 2018-08-10T04:20:54Z fix ---- --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org