If possible, find an existing meaningful benchmark that we can add stroom to. If not, create a benchmark that can also be run on hadoop, spark, mongo, postgres and so forth to do a comparative benchmark... the goal is not to show stroom beating anyone on runtime, but to show the development effort needed while showing what kind of jobs and data sizes stroom is "fast enough" for.
If possible, find an existing meaningful benchmark that we can add stroom to. If not, create a benchmark that can also be run on hadoop, spark, mongo, postgres and so forth to do a comparative benchmark... the goal is not to show stroom beating anyone on runtime, but to show the development effort needed while showing what kind of jobs and data sizes stroom is "fast enough" for.