Re: [infinispan-dev] MapReduce on REPL and Mapper behaviour

On 10.08.2012 13:36, Mircea Markus wrote: > >> >> 2) In case I do not have an Reduce job, I'd like to avoid provide >> Reducer. Creating feature request that reducedWith(...) is optional? >> > As you don't need a reducer, have you considered using the DistributedExecutorService with a DistributedCallable for that? I wanted to execute an operation only once on every key, independend how many nodes are owner of the entry. The DistCallable approach was my first solution, where I checked at every node for every entry, if the current node is owner, before I modified it. So using the Mapper was a much simpler solution.