Github user WangTaoTheTonic commented on the pull request:
https://github.com/apache/spark/pull/3607#issuecomment-67318029
I have tested in my cluster(3 NM nodes with 8GB memory each), with
configuration:
`spark.driver.memory=4G
spark.yarn.am.memoryOverhead=1024
spark.yarn.am.memory=1G`
Before this patch, yarn-client mode will use 3 container, the memory used
on NM nodes are 4.5G, 1G and 1G, and in yarn-cluster mode they are 5.5G and 1G.
After this patch, yarn-client mode use 1G and 3G, yarn-cluster mode use 1G
and 5.5G.
So the behaviour met our expects. @tgravescs @andrewor14 Is it ok to go?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
