HIVE-29367: prevent Long overflows in ConvertJoinMapJoin #6237
+475
−11
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,164 @@ | ||
| /* | ||
| * 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.hadoop.hive.ql.optimizer; | ||
|
|
||
| import static org.junit.jupiter.api.Assertions.assertEquals; | ||
| import static org.junit.jupiter.api.Assertions.assertNotNull; | ||
| import static org.junit.jupiter.api.Assertions.assertTrue; | ||
| import static org.mockito.Mockito.mock; | ||
| import static org.mockito.Mockito.when; | ||
|
|
||
| import java.lang.reflect.Method; | ||
| import java.util.ArrayList; | ||
| import java.util.Arrays; | ||
| import java.util.Collections; | ||
| import java.util.List; | ||
|
|
||
| import org.apache.hadoop.hive.ql.exec.Operator; | ||
| import org.apache.hadoop.hive.ql.plan.ColStatistics; | ||
| import org.apache.hadoop.hive.ql.plan.Statistics; | ||
| import org.junit.jupiter.api.Test; | ||
|
|
||
| class TestConvertJoinMapJoin { | ||
|
|
||
| @Test | ||
| void testComputeOnlineDataSizeGenericLargeDataSize() { | ||
| ConvertJoinMapJoin converter = new ConvertJoinMapJoin(); | ||
| converter.hashTableLoadFactor = 0.75f; | ||
| Statistics stats = new Statistics(1000L, Long.MAX_VALUE, 0L, 0L); | ||
|
|
||
| long result = converter.computeOnlineDataSizeGeneric(stats, 10L, 8L); | ||
|
|
||
| assertTrue(result >= 0, "Result should not be negative due to overflow"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeOnlineDataSizeGenericLargeNumRowsWithOverhead() { | ||
| ConvertJoinMapJoin converter = new ConvertJoinMapJoin(); | ||
| converter.hashTableLoadFactor = 0.75f; | ||
| Statistics stats = new Statistics(Long.MAX_VALUE / 2, 1000L, 0L, 0L); | ||
|
|
||
| long result = converter.computeOnlineDataSizeGeneric(stats, Long.MAX_VALUE / 4, Long.MAX_VALUE / 4); | ||
|
|
||
| assertTrue(result >= 0, "Result should not be negative due to overflow"); | ||
| assertEquals(Long.MAX_VALUE, result, "Result should saturate at Long.MAX_VALUE"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeOnlineDataSizeGenericNumNullsLargerThanNumRows() { | ||
| ConvertJoinMapJoin converter = new ConvertJoinMapJoin(); | ||
| converter.hashTableLoadFactor = 0.75f; | ||
| Statistics stats = new Statistics(100L, 10000L, 0L, 0L); | ||
| List<ColStatistics> colStats = new ArrayList<>(); | ||
| ColStatistics cs = new ColStatistics("col1", "string"); | ||
| cs.setNumNulls(Long.MAX_VALUE); | ||
| colStats.add(cs); | ||
| stats.setColumnStats(colStats); | ||
|
|
||
| long result = converter.computeOnlineDataSizeGeneric(stats, 10L, 8L); | ||
|
|
||
| assertTrue(result >= 0, "Result should not be negative due to underflow in nonNullCount"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeOnlineDataSizeGenericSmallDataSizeLargeAdjustment() { | ||
| ConvertJoinMapJoin converter = new ConvertJoinMapJoin(); | ||
| converter.hashTableLoadFactor = 0.75f; | ||
| Statistics stats = new Statistics(1000000L, 100L, 0L, 0L); | ||
| List<ColStatistics> colStats = new ArrayList<>(); | ||
| ColStatistics cs = new ColStatistics("col1", "string"); | ||
| cs.setNumNulls(0L); | ||
| colStats.add(cs); | ||
| stats.setColumnStats(colStats); | ||
|
|
||
| long result = converter.computeOnlineDataSizeGeneric(stats, 10L, 8L); | ||
|
|
||
| assertTrue(result >= 0, "Result should not be negative when adjustment > dataSize"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeOnlineDataSizeGenericAllExtremeValues() { | ||
| ConvertJoinMapJoin converter = new ConvertJoinMapJoin(); | ||
| converter.hashTableLoadFactor = 0.75f; | ||
| Statistics stats = new Statistics(Long.MAX_VALUE, Long.MAX_VALUE, 0L, 0L); | ||
| List<ColStatistics> colStats = new ArrayList<>(); | ||
| ColStatistics cs = new ColStatistics("col1", "string"); | ||
| cs.setNumNulls(Long.MAX_VALUE); | ||
| colStats.add(cs); | ||
| stats.setColumnStats(colStats); | ||
|
|
||
| long result = converter.computeOnlineDataSizeGeneric(stats, Long.MAX_VALUE, Long.MAX_VALUE); | ||
|
|
||
| assertTrue(result >= 0, "Result should not be negative with extreme values"); | ||
| assertEquals(Long.MAX_VALUE, result, "Result should saturate at Long.MAX_VALUE"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeCumulativeCardinalityWithParentsOverflow() { | ||
| Operator<?> parent1 = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(parent1.getParentOperators()).thenReturn(Collections.emptyList()); | ||
| Operator<?> parent2 = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(parent2.getParentOperators()).thenReturn(Collections.emptyList()); | ||
| Operator<?> mockOp = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(mockOp.getParentOperators()).thenReturn(Arrays.asList(parent1, parent2)); | ||
|
|
||
| Long result = invokeComputeCumulativeCardinality(mockOp); | ||
|
|
||
| assertNotNull(result, "Result should not be null"); | ||
| assertTrue(result >= 0, "Result should not be negative due to overflow"); | ||
| assertEquals(Long.MAX_VALUE, result.longValue(), "Result should saturate at Long.MAX_VALUE"); | ||
| } | ||
|
|
||
| @Test | ||
| void testComputeCumulativeCardinalityDeepTreeOverflow() { | ||
| Operator<?> leaf = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(leaf.getParentOperators()).thenReturn(Collections.emptyList()); | ||
| Operator<?> mid1 = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(mid1.getParentOperators()).thenReturn(Collections.singletonList(leaf)); | ||
| Operator<?> mid2 = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(mid2.getParentOperators()).thenReturn(Collections.singletonList(mid1)); | ||
| Operator<?> root = createMockOperatorWithStats(Long.MAX_VALUE / 2); | ||
| when(root.getParentOperators()).thenReturn(Collections.singletonList(mid2)); | ||
|
|
||
| Long result = invokeComputeCumulativeCardinality(root); | ||
|
|
||
| assertNotNull(result, "Result should not be null"); | ||
| assertTrue(result >= 0, "Result should not be negative due to overflow"); | ||
| assertEquals(Long.MAX_VALUE, result.longValue(), "Result should saturate at Long.MAX_VALUE"); | ||
| } | ||
|
|
||
| @SuppressWarnings("unchecked") | ||
| private Operator<?> createMockOperatorWithStats(long numRows) { | ||
| Operator<?> mockOp = mock(Operator.class); | ||
| Statistics stats = new Statistics(numRows, numRows * 100, 0L, 0L); | ||
| when(mockOp.getStatistics()).thenReturn(stats); | ||
| return mockOp; | ||
| } | ||
|
|
||
| private Long invokeComputeCumulativeCardinality(Operator<?> op) { | ||
| try { | ||
| Method method = ConvertJoinMapJoin.class.getDeclaredMethod( | ||
| "computeCumulativeCardinality", Operator.class); | ||
| method.setAccessible(true); | ||
| return (Long) method.invoke(null, op); | ||
| } catch (Exception e) { | ||
| throw new RuntimeException(e); | ||
| } | ||
| } | ||
| } |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,26 @@ | ||
| -- Test overflow handling in computeOnlineDataSize with Long.MAX_VALUE statistics | ||
|
|
||
| SET hive.auto.convert.join=true; | ||
| SET hive.auto.convert.join.noconditionaltask=true; | ||
| SET hive.auto.convert.join.noconditionaltask.size=10000000; | ||
|
|
||
| CREATE TABLE t1 (k BIGINT, v STRING); | ||
| CREATE TABLE t2 (k BIGINT, v STRING); | ||
|
|
||
| -- Case 1: Normal statistics - t1 fits in 10MB threshold, MapJoin expected | ||
| ALTER TABLE t1 UPDATE STATISTICS SET('numRows'='10000','rawDataSize'='100000'); | ||
| ALTER TABLE t1 UPDATE STATISTICS FOR COLUMN k SET('numDVs'='10000','numNulls'='0'); | ||
| ALTER TABLE t1 UPDATE STATISTICS FOR COLUMN v SET('numDVs'='10000','numNulls'='0','avgColLen'='10.0','maxColLen'='20'); | ||
|
|
||
| ALTER TABLE t2 UPDATE STATISTICS SET('numRows'='1000000','rawDataSize'='10000000'); | ||
| ALTER TABLE t2 UPDATE STATISTICS FOR COLUMN k SET('numDVs'='1000000','numNulls'='0'); | ||
| ALTER TABLE t2 UPDATE STATISTICS FOR COLUMN v SET('numDVs'='1000000','numNulls'='0','avgColLen'='10.0','maxColLen'='20'); | ||
|
|
||
| EXPLAIN SELECT t1.k, t2.v FROM t1 JOIN t2 ON t1.k = t2.k; | ||
|
|
||
| -- Case 2: Long.MAX_VALUE numRows - without fix, overflow causes negative size and incorrect MapJoin | ||
| ALTER TABLE t1 UPDATE STATISTICS SET('numRows'='9223372036854775807','rawDataSize'='9223372036854775807'); | ||
| ALTER TABLE t1 UPDATE STATISTICS FOR COLUMN k SET('numDVs'='1000','numNulls'='0'); | ||
| ALTER TABLE t1 UPDATE STATISTICS FOR COLUMN v SET('numDVs'='1000','numNulls'='0','avgColLen'='10.0','maxColLen'='20'); | ||
|
|
||
| EXPLAIN SELECT t1.k, t1.v, t2.v FROM t1 JOIN t2 ON t1.k = t2.k; |
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure it's appropriate to use safeAdd for a table size?
on the other side
hashTableDataSizeAdjustmentdoes that as well, so I guess it's finecc @zabetak, @thomasrebele
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it's fine. The total size here does not need to be 100% correct, it's just an estimation that influences the join decision. Might make sense to rename it to
estimatedTotalSize.