Akshat-Jain commented on PR #17170: URL: https://github.com/apache/druid/pull/17170#issuecomment-2537836764
> In addition to checking out the comments, please check the effect on query performance after applying this patch in a scenario where there are many (like 100) FireHydrants per Sink. @gianm I've added a benchmark for this. Reporting the data for the same below. For master branch code: ``` Benchmark (numFireHydrants) Mode Cnt Score Error Units SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 10 avgt 5 0.336 ± 0.014 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 50 avgt 5 1.937 ± 0.773 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 100 avgt 5 3.294 ± 1.997 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 200 avgt 5 5.931 ± 0.153 ms/op ``` With this PR's code changes: ``` Benchmark (numFireHydrants) Mode Cnt Score Error Units SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 10 avgt 5 0.329 ± 0.007 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 50 avgt 5 0.738 ± 0.164 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 100 avgt 5 1.324 ± 0.124 ms/op SinkQuerySegmentWalkerBenchmark.emitSinkMetrics 200 avgt 5 2.650 ± 0.454 ms/op ``` We can see the improvement in performance with this PR's code change compared to master code. Hope this works! -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
