deemoliu opened a new pull request, #11747: URL: https://github.com/apache/pinot/pull/11747
`enhancement`: Add pre-aggregation support for distinct count hll++ function. rohityadav1993 has done a CPU usage benchmark for hll++ functions. <meta charset="utf-8"><b style="font-weight:normal;" id="docs-internal-guid-1d745475-7fff-70c5-721e-738f40d2f8c6"><div dir="ltr" style="margin-left:0pt;" align="left"> | Query1 (% of CPU) | Query2 (% of CPU) -- | -- | -- Distinct CountdistinctCount(driver_uuid) | min: 1.54max: 5.79avg: 2.78 | min: 0.33max: 4.51avg: 1.15 HLL (default log2m=8)distinctCountHLL(driver_uuid) | min: 0.07max: 2.81avg: 0.68 | min: 0.46max: 1.38avg: 0.76 HLL log2m=9 (e=0.035)distinctCountHLL(driver_uuid, 9) | min: 0.07max: 1.62avg: 0.73 | min: 0.34max: 1.42avg: 0.76 HLL log2m=10 (e=0.03)distinctCountHLL(driver_uuid, 10) | min: 0.07max: 1.58avg: 0.88 | min: 0.52max: 1.27avg: 0.73 HLL++ default p=14distinctCountHLLPLUS(driver_uuid) | min: 0.06max: 1.99avg: 0.63 | min: 0.34max: 1.46avg: 0.75 hLL++ p=12distinctCountHLLPLUS(driver_uuid, 12) | min: 0.09max: 3.33avg: 0.35 | min: 0.31max: 1.57avg: 0.85 HLL++ p=10distinctCountHLLPLUS(driver_uuid, 12) | min: 0.08max: 4.41avg: 1.37 | min: 0.48max: 1.26avg: 0.70 </div></b> We want to enable hll++ in pre-aggregration to further perform some benchmarking with distinctCountHLL++ on pre-computed hll bytes fields. -- 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]
