Re: df.groupBy('m).agg(sum('n)).show dies with 10^3 elements?

2016-09-06 Thread Jacek Laskowski
Hi Josh, Yes, that seems to be the issue. As I commented out in the JIRA, just yesterday (after I had sent the email), such simple queries like the following killed spark-shell: Seq(1).toDF.groupBy('value).count.show Hoping to see it get resolved soon. If there's anything I could help you with t

Re: df.groupBy('m).agg(sum('n)).show dies with 10^3 elements?

2016-09-06 Thread Josh Rosen
I think that this is a simpler case of https://issues.apache.org/jira/browse/SPARK-17405. I'm going to comment on that ticket with your simpler reproduction. On Tue, Sep 6, 2016 at 1:32 PM Jacek Laskowski wrote: > Hi, > > I'm concerned with the OOME in local mode with the version built today: >

df.groupBy('m).agg(sum('n)).show dies with 10^3 elements?

2016-09-06 Thread Jacek Laskowski
Hi, I'm concerned with the OOME in local mode with the version built today: scala> val intsMM = 1 to math.pow(10, 3).toInt intsMM: scala.collection.immutable.Range.Inclusive = Range(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31,