zipkin-sparkstreaming
zipkin-sparkstreaming copied to clipboard
Document using with YARN
I'll preface this by saying I have zero operational knowledge about spark so...
In order to use this in AWS with EMR we will need to be able to connect to a YARN controlled Spark cluster. I'm not sure if currently the project even supports YARN but it would be good to know and have some documentation around if/when it does.
looks like normally yarn is a different distribution, which special cases the magic value "yarn" as the cluster name. It might be possible to hack our main class to do the same, or we document using spark-submit, treating this app as a normal jar.
here are some related links
https://spark.apache.org/docs/1.6.3/running-on-yarn.html http://blog.sequenceiq.com/blog/2014/08/22/spark-submit-in-java/
EDIT: jinx again.. 😄