I think these are great ideas for simplifying the getting started experience across runners -- we'd love a contribution in this space!
On Mon, Feb 20, 2017 at 12:46 AM, Jean-Baptiste Onofré <[email protected]> wrote: > Hi Nitin, > > It sounds like a good candidate for blog or documentation, or even an > example. > > I have a step-by-step example of a pipeline running on Spark with Mesos > (not yet Kubernetes) if you are interested. > > Regards > JB > > > On 02/20/2017 08:47 AM, Nitin Lamba wrote: > >> Hi, >> >> Trying to restart this thread from last November [1]. Packaging an >> end-2-end Beam example for k8s environment, similar to the one from the >> TensorFlow team [2], may be interesting to look at. The logical >> progression >> is: >> >> - Start with an example using the local (java) runner >> - Build the next one for Spark; k8s repo already has Spark v1.5.x as an >> example [3] that can be updated/ modified >> - Other runners to follow using Spark as a template >> >> Let me know if there is interest in pursuing/ collaborating on this. >> >> Thanks, >> Nitin >> >> [1] >> https://www.mail-archive.com/[email protected]/msg00881.html >> [2] https://tensorflow.github.io/serving/serving_inception >> [3] https://github.com/kubernetes/kubernetes/tree/master/examples/spark >> >> > -- > Jean-Baptiste Onofré > [email protected] > http://blog.nanthrax.net > Talend - http://www.talend.com >
