Alex, You're right about the parameter *receive.buffer.bytes*. When using his default value, my consumer throughput is about 3 000 msg / sec when my producer send about 5000 msg / sec.
When I setting the receive.buffer.bytes to a higher value, I have the same throughput than the producer. This is a good new for us and my team, many thanks for sharing your experience with reactive kafka ! In my case, I also put "fetch.max.wait.ms" to "1000". Le lundi 26 juin 2017 19:40:27 UTC+2, Alex Cozzi a écrit : > > Great! > about performance: I tried to tweak a lot of parameters, but what I found > has most influence on the throughput on the reading side seems to be > "receive.buffer.bytes". The optimum varies depending on your image size and > other factors, but you can try to do a bit of parameter search. In my case > this seems to be the best. Also, if it is acceptable to your application, > using autocommit speeds this up quite a bit, but you must accept a bit of > uncertainty as to when you commit an offset. > > receive.buffer.bytes = 131072 # (128 * 1024) > > -- >>>>>>>>>> Read the docs: http://akka.io/docs/ >>>>>>>>>> Check the FAQ: >>>>>>>>>> http://doc.akka.io/docs/akka/current/additional/faq.html >>>>>>>>>> Search the archives: https://groups.google.com/group/akka-user --- You received this message because you are subscribed to the Google Groups "Akka User List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/akka-user. For more options, visit https://groups.google.com/d/optout.
