Hi Johannes, Thanks for your question. You can try to implement the Value interface for your Vector PoJo. It has to have an empty constructor and implement the write and read methods of the interface for serialization.
Based on your description, I've implemented an example to demonstrate the use of the Value interface. It would be interesting to hear from you whether you could decrease the serialization time using this serialization method. Best regards, Max On Tue, Feb 24, 2015 at 11:13 AM, Kirschnick, Johannes <johannes.kirschn...@tu-berlin.de> wrote: > Hi list, > > > I have a general question on as to whether it's possible to significantly > speed up the processing by cutting down on the serialization costs during > iterations. > > > The basic setup that I have are a couple of vectors that are repeatedly > mutated (added & multiplied) as part of an iterative run within a reducer. > > A vector is basically "just" an array of doubles - all of the same size. > > > I noticed during simple profiling that roughly 50% of the execution time is > spent on serializing the data in using the > com.esotericsoftware.kryo.serializers.DefaultArraySerializers in Kryo. > > > I know that any custom operation would would varant custom processing, but > given the serialization contributes such a large amount of processing time to > the overall runtime it might very well be worthwhile > > > Is that currently exposed in any fashion to the user code, or are there some > hooks I could look into? > > > Thanks > > Johannes
