Is the first() being computed locally on the driver program? Maybe it's to hard to compute with the memory, etc available there. Take a look at the driver's log and see whether it has the message "Computing the requested partition locally".
Matei On Jul 22, 2014, at 12:04 PM, Nathan Kronenfeld <nkronenf...@oculusinfo.com> wrote: > I was wondering if anyone could provide an explanation for the behavior I'm > seeing. > > I have an RDD, call it foo, not too complex, with a maybe 8 level deep DAG > with 2 shuffles, not empty, not even terribly big - small enough that some > partitions could be empty. > > When I run foo.first, I get workers disconnecting, and applications die > When I run foo.mapPartitions.saveAsHadoopDataset, it works fine. > > Anyone got an explanation for why that might be? > > -Thanks, Nathan > > > -- > Nathan Kronenfeld > Senior Visualization Developer > Oculus Info Inc > 2 Berkeley Street, Suite 600, > Toronto, Ontario M5A 4J5 > Phone: +1-416-203-3003 x 238 > Email: nkronenf...@oculusinfo.com