ODIn/spark-instrumented-optimizer
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity
spark-instrumented-optimizer/sql/core/benchmarks
History
Max Gekk bef5828e12 [SPARK-31630][SQL] Fix perf regression by skipping timestamps rebasing after some threshold 
### What changes were proposed in this pull request?
Skip timestamps rebasing after a global threshold when there is no difference between Julian and Gregorian calendars. This allows to avoid checking hash maps of switch points, and fixes perf regressions in `toJavaTimestamp()` and `fromJavaTimestamp()`.

### Why are the changes needed?
The changes fix perf regressions of conversions to/from external type `java.sql.Timestamp`.

Before (see the PR's results https://github.com/apache/spark/pull/28440):
```
================================================================================================
Conversion from/to external types
================================================================================================

OpenJDK 64-Bit Server VM 1.8.0_252-8u252-b09-1~18.04-b09 on Linux 4.15.0-1063-aws
Intel(R) Xeon(R) CPU E5-2670 v2  2.50GHz
To/from Java's date-time:                 Best Time(ms)   Avg Time(ms)   Stdev(ms)    Rate(M/s)   Per Row(ns)   Relative
------------------------------------------------------------------------------------------------------------------------
From java.sql.Timestamp                             376            388          10         13.3          75.2       1.1X
Collect java.sql.Timestamp                         1878           1937          64          2.7         375.6       0.2X
```

After:
```
================================================================================================
Conversion from/to external types
================================================================================================

OpenJDK 64-Bit Server VM 1.8.0_252-8u252-b09-1~18.04-b09 on Linux 4.15.0-1063-aws
Intel(R) Xeon(R) CPU E5-2670 v2  2.50GHz
To/from Java's date-time:                 Best Time(ms)   Avg Time(ms)   Stdev(ms)    Rate(M/s)   Per Row(ns)   Relative
------------------------------------------------------------------------------------------------------------------------
From java.sql.Timestamp                             249            264          24         20.1          49.8       1.7X
Collect java.sql.Timestamp                         1503           1523          24          3.3         300.5       0.3X
```

Perf improvements in average of:

1. From java.sql.Timestamp is ~ 34%
2. To java.sql.Timestamps is ~16%

### Does this PR introduce _any_ user-facing change?
No

### How was this patch tested?
By existing test suites `DateTimeUtilsSuite` and `RebaseDateTimeSuite`.

Closes #28441 from MaxGekk/opt-rebase-common-threshold.

Authored-by: Max Gekk <max.gekk@gmail.com>
Signed-off-by: Wenchen Fan <wenchen@databricks.com>
2020-05-05 14:11:53 +00:00
..
AggregateBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
AggregateBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
BloomFilterBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
BloomFilterBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
BuiltInDataSourceWriteBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
BuiltInDataSourceWriteBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
ColumnarBatchBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
ColumnarBatchBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
CompressionSchemeBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
CompressionSchemeBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
CSVBenchmark-jdk11-results.txt [SPARK-31414][SQL] Fix performance regression with new TimestampFormatter for json and csv time parsing 2020-04-13 03:11:28 +00:00
CSVBenchmark-results.txt [SPARK-31414][SQL] Fix performance regression with new TimestampFormatter for json and csv time parsing 2020-04-13 03:11:28 +00:00
DatasetBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
DatasetBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
DataSourceReadBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
DataSourceReadBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
DateTimeBenchmark-jdk11-results.txt [SPARK-31630][SQL] Fix perf regression by skipping timestamps rebasing after some threshold 2020-05-05 14:11:53 +00:00
DateTimeBenchmark-results.txt [SPARK-31630][SQL] Fix perf regression by skipping timestamps rebasing after some threshold 2020-05-05 14:11:53 +00:00
DateTimeRebaseBenchmark-jdk11-results.txt [SPARK-31630][SQL] Fix perf regression by skipping timestamps rebasing after some threshold 2020-05-05 14:11:53 +00:00
DateTimeRebaseBenchmark-results.txt [SPARK-31630][SQL] Fix perf regression by skipping timestamps rebasing after some threshold 2020-05-05 14:11:53 +00:00
ExternalAppendOnlyUnsafeRowArrayBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
ExternalAppendOnlyUnsafeRowArrayBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
ExtractBenchmark-jdk11-results.txt [SPARK-31507][SQL] Remove uncommon fields support and update some fields with meaningful names for extract function 2020-04-22 10:24:49 +00:00
ExtractBenchmark-results.txt [SPARK-31507][SQL] Remove uncommon fields support and update some fields with meaningful names for extract function 2020-04-22 10:24:49 +00:00
FilterPushdownBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
FilterPushdownBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
HashedRelationMetricsBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
HashedRelationMetricsBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
InExpressionBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
InExpressionBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
IntervalBenchmark-jdk11-results.txt [SPARK-31129][SQL][TESTS] Fix IntervalBenchmark and DateTimeBenchmark 2020-03-12 12:59:29 -07:00
IntervalBenchmark-results.txt [SPARK-31129][SQL][TESTS] Fix IntervalBenchmark and DateTimeBenchmark 2020-03-12 12:59:29 -07:00
JoinBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
JoinBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
JsonBenchmark-jdk11-results.txt [SPARK-31414][SQL] Fix performance regression with new TimestampFormatter for json and csv time parsing 2020-04-13 03:11:28 +00:00
JsonBenchmark-results.txt [SPARK-31414][SQL] Fix performance regression with new TimestampFormatter for json and csv time parsing 2020-04-13 03:11:28 +00:00
MakeDateTimeBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
MakeDateTimeBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
MetricsAggregationBenchmark-jdk11-results.txt [SPARK-29562][SQL] Speed up and slim down metric aggregation in SQL listener 2019-10-24 22:18:10 -07:00
MetricsAggregationBenchmark-results.txt [SPARK-29562][SQL] Speed up and slim down metric aggregation in SQL listener 2019-10-24 22:18:10 -07:00
MiscBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
MiscBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
OrcNestedSchemaPruningBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
OrcNestedSchemaPruningBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
OrcV2NestedSchemaPruningBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
OrcV2NestedSchemaPruningBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
ParquetNestedPredicatePushDownBenchmark-jdk11-results.txt [SPARK-31364][SQL][TESTS] Benchmark Parquet Nested Field Predicate Pushdown 2020-04-24 22:10:58 +00:00
ParquetNestedPredicatePushDownBenchmark-results.txt [SPARK-31364][SQL][TESTS] Benchmark Parquet Nested Field Predicate Pushdown 2020-04-24 22:10:58 +00:00
ParquetNestedSchemaPruningBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
ParquetNestedSchemaPruningBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
PrimitiveArrayBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
PrimitiveArrayBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
RangeBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
RangeBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
SortBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
SortBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
TPCDSQueryBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
TPCDSQueryBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
UDFBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
UDFBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
UnsafeArrayDataBenchmark-jdk11-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
UnsafeArrayDataBenchmark-results.txt [SPARK-29320][TESTS] Compare sql/core module in JDK8/11 (Part 1) 2019-10-03 08:58:25 -07:00
WideSchemaBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
WideSchemaBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
WideTableBenchmark-jdk11-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00
WideTableBenchmark-results.txt [SPARK-30409][SPARK-29173][SQL][TESTS] Use NoOp datasource in SQL benchmarks 2020-01-12 13:18:19 -08:00