You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@druid.apache.org by GitBox <gi...@apache.org> on 2021/08/25 04:06:55 UTC
[GitHub] [druid] jihoonson commented on a change in pull request #11574: Configurable maxStreamLength for doubles sketches
jihoonson commented on a change in pull request #11574:
URL: https://github.com/apache/druid/pull/11574#discussion_r695375561
##########
File path: extensions-core/datasketches/src/main/java/org/apache/druid/query/aggregation/datasketches/quantiles/DoublesSketches.java
##########
@@ -0,0 +1,53 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements. See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership. The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License. You may obtain a copy of the License at
+ *
+ * http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied. See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+
+package org.apache.druid.query.aggregation.datasketches.quantiles;
+
+import org.apache.druid.java.util.common.ISE;
+
+public final class DoublesSketches
+{
+ /**
+ * Runs the given task that updates a Doubles sketch backed by a direct byte buffer. This method intentionally
+ * accepts the update task as a {@link Runnable} instead of accpeting parameters of a sketch and other values
+ * needed for the update. This is to avoid any potential performance impact especially when the sketch is updated
Review comment:
I ran the benchmark I added in https://github.com/apache/druid/pull/11574/commits/7be41051242f16db225dfce48f86adc9b1966c3a. It doesn't seem to have any difference.
```
// 20: doubles sketches
query: "SELECT APPROX_QUANTILE_DS(sumFloatNormal, 0.5), DS_QUANTILES_SKETCH(maxLongUniform) FROM foo"
master
Benchmark (query) (rowsPerSegment) (vectorize) Mode Cnt Score Error Units
SqlBenchmark.querySql 20 5000000 false avgt 15 341.659 ± 0.932 ms/op
SqlBenchmark.querySql 20 5000000 force avgt 15 446.572 ± 3.279 ms/op
PR
Benchmark (query) (rowsPerSegment) (vectorize) Mode Cnt Score Error Units
SqlBenchmark.querySql 20 5000000 false avgt 15 341.956 ± 2.478 ms/op
SqlBenchmark.querySql 20 5000000 force avgt 15 442.048 ± 1.420 ms/op
```
Interestingly, the query was _slower_ when the vectorization was on. AFAIT, the difference is that the non-vectorized version uses `HeapUpdateDoublesSketch` while the vectorized one uses `DirectUpdateDoublesSketch`. Attaching flame graphs.
Non-vectorized
Vectorized
--
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: commits-unsubscribe@druid.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For additional commands, e-mail: commits-help@druid.apache.org