Hi Lydia, what do you mean with master? Usually when you submit a program to the cluster and don’t specify the parallelism in your program, then it will be executed with the parallelism.default value as parallelism. You can specify the value in your cluster configuration flink-config.yaml file. Alternatively you can always specify the parallelism via the CLI client with the -p option.
Cheers, Till On Thu, Jan 28, 2016 at 9:53 AM, Lydia Ickler <ickle...@googlemail.com> wrote: > Hi all, > > I am doing some operations on a DataSet<Tuple3<Integer,Integer,Double>> … > (see code below) > When I run my program on a cluster with 3 machines I can see within the > web client that only my master is executing the program. > Do I have to specify somewhere that all machines have to participate? > Usually the cluster executes in parallel. > > Any suggestions? > > Best regards, > Lydia > > DataSet<Tuple3<Integer, Integer, Double>> matrixA = readMatrix(env, input); > > DataSet<Tuple3<Integer, Integer, Double>> initial = matrixA.groupBy(0).sum(2); > > //normalize by maximum value > initial = initial.cross(initial.max(2)).map(new normalizeByMax()); > > matrixA.join(initial).where(1).equalTo(0) > > .map(new ProjectJoinResultMapper()).groupBy(0, 1).sum(2); > > >
