[
https://issues.apache.org/jira/browse/BEAM-982?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15883673#comment-15883673
]
Amit Sela commented on BEAM-982:
--------------------------------
[~jbonofre] {{spark.serializer}} is already hard-coded into the runner, and
takes highest precedence.
I don't believe {{spark.kryoserializer.buffer.max}} was related specifically to
your problem.
> Document spark configuration update
> -----------------------------------
>
> Key: BEAM-982
> URL: https://issues.apache.org/jira/browse/BEAM-982
> Project: Beam
> Issue Type: Improvement
> Components: runner-spark
> Reporter: Jean-Baptiste Onofré
> Priority: Minor
>
> When using {{spark-submit}} to submit a pipeline to a spark cluster, I had to
> change the {{conf/spark-defaults.conf}} with the following:
> {code}
> spark.serializer org.apache.spark.serializer.KryoSerializer
> spark.kryoserializer.buffer.max 1024
> {code}
> I think it would make sense to add a note in the Spark README or website
> about this.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
