StefanRRichter commented on a change in pull request #8409: [FLINK-12478]
Decompose monolithic run-loops in StreamTask implementa…
URL: https://github.com/apache/flink/pull/8409#discussion_r284663084
##########
File path:
flink-streaming-java/src/main/java/org/apache/flink/streaming/runtime/tasks/OneInputStreamTask.java
##########
@@ -98,13 +96,8 @@ public void init() throws Exception {
}
@Override
- protected void run() throws Exception {
- // cache processor reference on the stack, to make the code
more JIT friendly
- final StreamInputProcessor<IN> inputProcessor =
this.inputProcessor;
-
- while (running && inputProcessor.processInput()) {
- // all the work happens in the "processInput" method
- }
+ protected boolean performDefaultAction() throws Exception {
Review comment:
I ran the following:
```
env.setParallelism(1);
source
.map(new MultiplyByTwo())
.addSink(new DiscardingSink<>());
master:
Benchmark Mode Cnt Score Error Units
SumLongsBenchmark.benchmarkCount thrpt 30 16240.866 ± 311.845 ops/ms
master + #8442:
Benchmark Mode Cnt Score Error Units
SumLongsBenchmark.benchmarkCount thrpt 30 16540.537 ± 295.100 ops/ms
```
```
env.setParallelism(1);
source
.map(new MultiplyByTwo())
.rebalance()
.map((x) -> x)
.addSink(new DiscardingSink<>());
master:
Benchmark Mode Cnt Score Error Units
SumLongsBenchmark.benchmarkCount thrpt 30 9715.334 ± 170.632 ops/ms
master + #8442:
Benchmark Mode Cnt Score Error Units
SumLongsBenchmark.benchmarkCount thrpt 30 9932.752 ± 124.494 ops/ms
```
I would conclude the changes run as fast or a notch faster than master on my
local machine.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
