Sure, so there would be one flux file defining the topology that all the parsers use. Each parser topology has a:
- Kafka Spout - Parser bolt - Error Writer Bolt - Invalid Message Writer Bolt The actual parser that you are using and the writer that you write out to from the parser bolt is parameterized inside of the zookeeper config for that parser. Since the actual parser implementation and writer to use are already parameterized inside of the config, we should be able to get away with one flux file and a property that indicates the name of the config in zookeeper. Casey On Mon, Sep 26, 2016 at 1:19 PM, Otto Fowler <[email protected]> wrote: > Casey, can you explain the one flux file with separate configurations a > little more? > > On September 26, 2016 at 12:22:45, Casey Stella ([email protected]) > wrote: > > We moved the parser topologies away from Flux early on, mostly because the > way it was working originally would require separate flux files per > topology. I think that we are, with the current architecture, would be > able to use one flux file for all of the topology (different params, > obviously). Flux is quite configurable and if we could use one flux file > and just have separate configs per parser topology, then I think we might > get the best of both worlds. > > What do you guys think? > >
