Hi Kevin, I saw #235, it changes too much to catch at a glance. Do you think It's a kind of solution that I can make a new interpreter with my own setting? I'll try to that, but first of all, I use and have a spark 1.2 cluster only. Could you tell me your schedule for merging spark_1.2 branch into master?
And the purpose of spark.* is tuning spark as well. When I use a mesos cluster, spark.exeutor.uri is needed for specific spark executor. and Kyro is also needed and compression, logging event, and so on. I think It's a good idea to support spark-defaults.conf and spark-env.sh. How about you? Regards, JL On Fri, Jan 30, 2015 at 12:41 AM, Kevin Kim (Sangwoo) <[email protected]> wrote: > Great to hear you found the way! > Maybe we can find out a better way than the workaround in a near future. > > And what is the purpose of setting spark.*? > Can you be more specific? > > [I guess you can set spark.* via #235] > -- 이종열, Jongyoul Lee, 李宗烈 http://madeng.net
