pnowojski commented on pull request #23: URL: https://github.com/apache/flink-benchmarks/pull/23#issuecomment-884219160
1. It looks like [the place you mentioned](https://github.com/apache/flink/tree/master/flink-state-backends/flink-statebackend-rocksdb/src/test/java/org/apache/flink/contrib/streaming/state/benchmark) is a good one to start About the 30 minutes run. Yes, dropping `heap` and picking a single value for `subTaskIndex` would help a lot to bring this time down. Also keep in mind the recommendation from the readme: > Rule of thumb is that you should increase the number of processed records up to a point where results stop improving visibly, while trying to keep the single benchmark invocation around/under 1 second. With single iteration 1 second, the total time should get down to ~4minutes. But there can be a trade off with too short single invocation, the benchmarks can be unstable. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
