from Hacker News

Concord – High Performance Stream Processing with C++ and Mesos

by karamazov on 5/13/16, 4:23 PM with 26 comments

  • by cstivers1978 on 5/13/16, 8:09 PM

    I really like the idea behind Concord. No longer have to stress about Hadoop/YARN platform, and I can use the language of my choice (haven't used JVM-based stack in ages).

    Is there documentation on adding more input/output sources?

  • by cstivers1978 on 5/13/16, 8:12 PM

    Is it possible to abstract the metadata store, which is currently Zookeeper, to be something else like etcd and consul?
  • by amn0408 on 5/13/16, 4:37 PM

    We use this at my company. Fast implementation even for those new to stream processing. Up and running in hours.
  • by followtherhythm on 5/13/16, 4:35 PM

    Hello, I am a software engineer at Concord. Happy to answer anyones questions
  • by LilBibby2342 on 5/13/16, 5:20 PM

    Is Storm still iterated and improved upon regularly? Curious if this is simpler than Storm because it's the single focus of the founders. Could be a big advantage of Concord, right?
  • by dangoldin on 5/13/16, 5:45 PM

    We've been messing around and looks very promising. It's much faster than Storm so at some point we'd love to move to it. Anyone knows how it compares against Flink?
  • by adev3000 on 5/13/16, 5:22 PM

    Awesome project and team. Highly recommended for predictable highspeed throughput. Faster than any other OS distributed processing engine.
  • by gratner on 5/13/16, 4:43 PM

    How does the performance compare to the traditional stream processors?
  • by tengudoll on 5/13/16, 4:42 PM

    10X faster than Storm. How? What are the tradeoffs?
  • by florianleibert on 5/13/16, 6:33 PM

    Super impressive performance!
  • by sonalmane on 5/13/16, 6:51 PM

    Great project and team!
  • by pastaking on 5/13/16, 4:40 PM

    Cool project!