Thanks Val. we have a c++ computing engine that currently works as an independent node today. We need to reuse this engine, and deploy it as a closure / callable across the cluster of nodes so that instances of this engine could work in parallel for various inputs, and once they all complete, there should be a reducer node that could join the results with certain rules and give out the results.
This use case closely matches your compute grid capability with fork/join / mapper/reducer. Since compute grid is not available in C++ today, we were thinking we we could have java API wrapping the closures using JNA for now and have the java compute grid do the rest. Are we thinking right? Will the load balancing work with this design? Please advice. Thanks, Arthi -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Compute-Grid-API-in-C-Scala-tp4456p4722.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.
