Hi Val, thank you getting back. The source type will be different, and the client will deal with the difference in a preprocess that's unique to that data source. Once that's done, it'll be ready for a standard transformation. So we can encapsulate it in a single service or callable.
Thanks for explaining the difference between client and servers. It clarifies things and make sense. We wanted to build different type of clients to accommodate the differences in the data sources. Does this make sense? what kind of solutions would you suggest? Jessie On Fri, Aug 18, 2017 at 3:34 PM, vkulichenko <valentin.kuliche...@gmail.com> wrote: > If data sources are different, then transformations are different as well, > no? If so, how are you going to encapsulate them in a single service? I > would really appreciate a concrete example, it might make this much easier. > I'm probably missing something now. > > Difference between clients and servers is the same is in other systems. > Servers store data, run services and computations, etc. while clients are > only used to interact with the cluster. > > -Val > > > > -- > View this message in context: http://apache-ignite-users. > 70518.x6.nabble.com/Streaming-test-tp14039p16310.html > Sent from the Apache Ignite Users mailing list archive at Nabble.com. >
