pvargacl commented on pull request #1230: URL: https://github.com/apache/hive/pull/1230#issuecomment-656531478
Benchmark results below. The 6000 sample is the one Rajesh provided, the others have 300 open/aborted txns. After the first run i modified the code, to not use ranges if the range < 5, this way is more effective. So i have cases for: - one single range - singles - ranges with length 2 - ranges with length 5 - many large ranges Benchmark (txnString) (useRange) Mode Cnt Score Error Units ValidReadTxnListBenchmark.writeAndParse MANY_RANGE_6000 true avgt 15 95.017 ± 6.665 us/op ValidReadTxnListBenchmark.writeAndParse MANY_RANGE_6000 false avgt 15 742.845 ± 40.087 us/op ValidReadTxnListBenchmark.writeAndParse RANGE_300 true avgt 15 3.475 ± 0.243 us/op ValidReadTxnListBenchmark.writeAndParse RANGE_300 false avgt 15 16.132 ± 0.422 us/op ValidReadTxnListBenchmark.writeAndParse SINGLES_300 true avgt 15 18.861 ± 0.712 us/op ValidReadTxnListBenchmark.writeAndParse SINGLES_300 false avgt 15 20.286 ± 1.183 us/op ValidReadTxnListBenchmark.writeAndParse DOUBLES_300 true avgt 15 19.415 ± 0.902 us/op ValidReadTxnListBenchmark.writeAndParse DOUBLES_300 false avgt 15 17.444 ± 0.298 us/op ValidReadTxnListBenchmark.writeAndParse FIVES_300 true avgt 15 15.242 ± 0.247 us/op ValidReadTxnListBenchmark.writeAndParse FIVES_300 false avgt 15 17.342 ± 0.475 us/op ---------------------------------------------------------------- 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: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org For additional commands, e-mail: gitbox-h...@hive.apache.org
