rx java - Usage of Apache Flink and RxJava -


i'm using apache flink , using rxjava inside of it, questions is: using both of them appropriate? because flink operations map functions , inside of them use rx intensively, take tuples flink , make asynchronous operations them (go db, write queue , on), don't ended using of methods flink expose me , programs steps returns json when rxjava ends processing.

can tell me if correct usage of flink or if there aa better way of need do. (like use pass observables between flink steps or that).

thanks.

flink has powerful tools handle state (e.g. in windows) [1,2], makes possible avoid calls other systems. example, instead of handling state in external key-value store, use (checkpointed) keyvaluestate in flink. handling state inside of flink more efficient/faster calling external systems/databases.

a problem asynchronous calls in flink programs can flink faster called systems, result in increasing number of open futures, leading memory problems. so, think, suggested use synchronous calls inside of flink operators slow down flink speed of called systems avoid memory leaks.

the last point has been discussed on mailing list: http://mail-archives.apache.org/mod_mbox/flink-user/201606.mbox/%3ccalrnvjvw=iu7pb76vmkk+bd6rdxfsdwxpmrtzk8=l5d8xe-o8q@mail.gmail.com%3e

[1] https://ci.apache.org/projects/flink/flink-docs-release-1.0/apis/streaming/windows.html [2]https://ci.apache.org/projects/flink/flink-docs-release-1.0/apis/streaming/state.html


Comments

Popular posts from this blog

sequelize.js - Sequelize group by with association includes id -

android - Robolectric "INTERNET permission is required" -

java - Android raising EPERM (Operation not permitted) when attempting to send UDP packet after network connection -