You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@spark.apache.org by Rex Xiong <by...@gmail.com> on 2015/10/10 07:53:44 UTC

Jar is cached in yarn-cluster mode?

I use "spark-submit -master yarn-cluster hdfs://.../a.jar ......" to submit
my app to yarn.
Then I update this a.jar in HDFS, run the command again, I found a line of
log that was been removed still exist in "yarn logs ....".
Is there a cache mechanism I need to disable?

Thanks