lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-3095] abstract partition filter logic to enable code reuse (#4454)

Browse Source 
* [HUDI-3095] abstract partition filter logic to enable code reuse

* [HUDI-3095] address reviews
This commit is contained in:
Yuwei XIAO
2021-12-31 13:37:52 +08:00
committed by GitHub
parent e88b5fd450
commit 2444f40a4b
16 changed files with 368 additions and 272 deletions
Download Patch File Download Diff File Expand all files Collapse all files

63
hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/client/clustering/plan/strategy/SparkRecentDaysClusteringPlanStrategy.java
View File

@@ -1,63 +0,0 @@
/*
* 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.hudi.client.clustering.plan.strategy;
import org.apache.hudi.client.common.HoodieSparkEngineContext;
import org.apache.hudi.common.model.HoodieRecordPayload;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.table.HoodieSparkCopyOnWriteTable;
import org.apache.hudi.table.HoodieSparkMergeOnReadTable;
import org.apache.log4j.LogManager;
import org.apache.log4j.Logger;
import java.util.Comparator;
import java.util.List;
import java.util.stream.Collectors;
/**
* Clustering Strategy that only looks at latest 'daybased.lookback.partitions' partitions.
*/
public class SparkRecentDaysClusteringPlanStrategy<T extends HoodieRecordPayload<T>>
extends SparkSizeBasedClusteringPlanStrategy<T> {
private static final Logger LOG = LogManager.getLogger(SparkRecentDaysClusteringPlanStrategy.class);
public SparkRecentDaysClusteringPlanStrategy(HoodieSparkCopyOnWriteTable<T> table,
HoodieSparkEngineContext engineContext,
HoodieWriteConfig writeConfig) {
super(table, engineContext, writeConfig);
}
public SparkRecentDaysClusteringPlanStrategy(HoodieSparkMergeOnReadTable<T> table,
HoodieSparkEngineContext engineContext,
HoodieWriteConfig writeConfig) {
super(table, engineContext, writeConfig);
}
@Override
protected List<String> filterPartitionPaths(List<String> partitionPaths) {
int targetPartitionsForClustering = getWriteConfig().getTargetPartitionsForClustering();
int skipPartitionsFromLatestForClustering = getWriteConfig().getSkipPartitionsFromLatestForClustering();
return partitionPaths.stream()
.sorted(Comparator.reverseOrder())
.skip(Math.max(skipPartitionsFromLatestForClustering, 0))
.limit(targetPartitionsForClustering > 0 ? targetPartitionsForClustering : partitionPaths.size())
.collect(Collectors.toList());
}
}

66
hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/client/clustering/plan/strategy/SparkSelectedPartitionsClusteringPlanStrategy.java
View File

@@ -1,66 +0,0 @@
/*
* 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.hudi.client.clustering.plan.strategy;
import org.apache.hudi.client.common.HoodieSparkEngineContext;
import org.apache.hudi.common.model.HoodieRecordPayload;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.table.HoodieSparkCopyOnWriteTable;
import org.apache.hudi.table.HoodieSparkMergeOnReadTable;
import org.apache.log4j.LogManager;
import org.apache.log4j.Logger;
import java.util.List;
import java.util.stream.Collectors;
import static org.apache.hudi.config.HoodieClusteringConfig.CLUSTERING_STRATEGY_PARAM_PREFIX;
/**
* Clustering Strategy to filter just specified partitions from [begin, end]. Note both begin and end are inclusive.
*/
public class SparkSelectedPartitionsClusteringPlanStrategy<T extends HoodieRecordPayload<T>>
extends SparkSizeBasedClusteringPlanStrategy<T> {
private static final Logger LOG = LogManager.getLogger(SparkSelectedPartitionsClusteringPlanStrategy.class);
public static final String CONF_BEGIN_PARTITION = CLUSTERING_STRATEGY_PARAM_PREFIX + "cluster.begin.partition";
public static final String CONF_END_PARTITION = CLUSTERING_STRATEGY_PARAM_PREFIX + "cluster.end.partition";
public SparkSelectedPartitionsClusteringPlanStrategy(HoodieSparkCopyOnWriteTable<T> table,
HoodieSparkEngineContext engineContext,
HoodieWriteConfig writeConfig) {
super(table, engineContext, writeConfig);
}
public SparkSelectedPartitionsClusteringPlanStrategy(HoodieSparkMergeOnReadTable<T> table,
HoodieSparkEngineContext engineContext,
HoodieWriteConfig writeConfig) {
super(table, engineContext, writeConfig);
}
@Override
protected List<String> filterPartitionPaths(List<String> partitionPaths) {
String beginPartition = getWriteConfig().getProps().getProperty(CONF_BEGIN_PARTITION);
String endPartition = getWriteConfig().getProps().getProperty(CONF_END_PARTITION);
List<String> filteredPartitions = partitionPaths.stream()
.filter(path -> path.compareTo(beginPartition) >= 0 && path.compareTo(endPartition) <= 0)
.collect(Collectors.toList());
LOG.info("Filtered to the following partitions: " + filteredPartitions);
return filteredPartitions;
}
}

1
hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/client/clustering/plan/strategy/SparkSingleFileSortPlanStrategy.java
View File

@@ -48,6 +48,7 @@ public class SparkSingleFileSortPlanStrategy<T extends HoodieRecordPayload<T>>
super(table, engineContext, writeConfig);
}
@Override
protected Stream<HoodieClusteringGroup> buildClusteringGroupsForPartition(String partitionPath, List<FileSlice> fileSlices) {
List<Pair<List<FileSlice>, Integer>> fileSliceGroups = fileSlices.stream()
.map(fileSlice -> Pair.of(Collections.singletonList(fileSlice), 1)).collect(Collectors.toList());

5
hudi-client/hudi-spark-client/src/main/java/org/apache/hudi/client/clustering/plan/strategy/SparkSizeBasedClusteringPlanStrategy.java
View File

@@ -115,11 +115,6 @@ public class SparkSizeBasedClusteringPlanStrategy<T extends HoodieRecordPayload<
return params;
}
@Override
protected List<String> filterPartitionPaths(List<String> partitionPaths) {
return partitionPaths;
}
@Override
protected Stream<FileSlice> getFileSlicesEligibleForClustering(final String partition) {
return super.getFileSlicesEligibleForClustering(partition)

66
hudi-client/hudi-spark-client/src/test/java/org/apache/hudi/client/clustering/plan/strategy/TestSparkRecentDaysClusteringPlanStrategy.java
View File

@@ -1,66 +0,0 @@
/*
* 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.hudi.client.clustering.plan.strategy;
import org.apache.hudi.client.common.HoodieSparkEngineContext;
import org.apache.hudi.config.HoodieClusteringConfig;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.table.HoodieSparkCopyOnWriteTable;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import org.mockito.Mock;
import java.util.ArrayList;
import java.util.List;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertSame;
public class TestSparkRecentDaysClusteringPlanStrategy {
@Mock
HoodieSparkCopyOnWriteTable table;
@Mock
HoodieSparkEngineContext context;
HoodieWriteConfig hoodieWriteConfig;
@BeforeEach
public void setUp() {
this.hoodieWriteConfig = HoodieWriteConfig
.newBuilder()
.withPath("Fake_Table_Path")
.withClusteringConfig(HoodieClusteringConfig
.newBuilder()
.withClusteringSkipPartitionsFromLatest(1)
.withClusteringTargetPartitions(1)
.build())
.build();
}
@Test
public void testFilterPartitionPaths() {
SparkRecentDaysClusteringPlanStrategy sg = new SparkRecentDaysClusteringPlanStrategy(table, context, hoodieWriteConfig);
ArrayList<String> fakeTimeBasedPartitionsPath = new ArrayList<>();
fakeTimeBasedPartitionsPath.add("20210718");
fakeTimeBasedPartitionsPath.add("20210716");
fakeTimeBasedPartitionsPath.add("20210719");
List list = sg.filterPartitionPaths(fakeTimeBasedPartitionsPath);
assertEquals(1, list.size());
assertSame("20210718", list.get(0));
}
}

107
hudi-client/hudi-spark-client/src/test/java/org/apache/hudi/table/action/cluster/strategy/TestSparkClusteringPlanPartitionFilter.java Normal file
View File

@@ -0,0 +1,107 @@
/*
* 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.hudi.table.action.cluster.strategy;
import org.apache.hudi.client.clustering.plan.strategy.SparkSizeBasedClusteringPlanStrategy;
import org.apache.hudi.client.common.HoodieSparkEngineContext;
import org.apache.hudi.config.HoodieClusteringConfig;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.table.HoodieSparkCopyOnWriteTable;
import org.apache.hudi.table.action.cluster.ClusteringPlanPartitionFilterMode;
import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test;
import org.mockito.Mock;
import java.util.ArrayList;
import java.util.List;
import static org.junit.jupiter.api.Assertions.assertEquals;
import static org.junit.jupiter.api.Assertions.assertSame;
public class TestSparkClusteringPlanPartitionFilter {
@Mock
HoodieSparkCopyOnWriteTable table;
@Mock
HoodieSparkEngineContext context;
HoodieWriteConfig.Builder hoodieWriteConfigBuilder;
@BeforeEach
public void setUp() {
this.hoodieWriteConfigBuilder = HoodieWriteConfig
.newBuilder()
.withPath("Fake_Table_Path");
}
@Test
public void testFilterPartitionNoFilter() {
HoodieWriteConfig config = hoodieWriteConfigBuilder.withClusteringConfig(HoodieClusteringConfig.newBuilder()
.withClusteringPlanPartitionFilterMode(ClusteringPlanPartitionFilterMode.NONE)
.build())
.build();
PartitionAwareClusteringPlanStrategy sg = new SparkSizeBasedClusteringPlanStrategy(table, context, config);
ArrayList<String> fakeTimeBasedPartitionsPath = new ArrayList<>();
fakeTimeBasedPartitionsPath.add("20210718");
fakeTimeBasedPartitionsPath.add("20210716");
fakeTimeBasedPartitionsPath.add("20210719");
List list = sg.filterPartitionPaths(fakeTimeBasedPartitionsPath);
assertEquals(3, list.size());
}
@Test
public void testFilterPartitionRecentDays() {
HoodieWriteConfig config = hoodieWriteConfigBuilder.withClusteringConfig(HoodieClusteringConfig.newBuilder()
.withClusteringSkipPartitionsFromLatest(1)
.withClusteringTargetPartitions(1)
.withClusteringPlanPartitionFilterMode(ClusteringPlanPartitionFilterMode.RECENT_DAYS)
.build())
.build();
PartitionAwareClusteringPlanStrategy sg = new SparkSizeBasedClusteringPlanStrategy(table, context, config);
ArrayList<String> fakeTimeBasedPartitionsPath = new ArrayList<>();
fakeTimeBasedPartitionsPath.add("20210718");
fakeTimeBasedPartitionsPath.add("20210716");
fakeTimeBasedPartitionsPath.add("20210719");
List list = sg.filterPartitionPaths(fakeTimeBasedPartitionsPath);
assertEquals(1, list.size());
assertSame("20210718", list.get(0));
}
@Test
public void testFilterPartitionSelectedPartitions() {
HoodieWriteConfig config = hoodieWriteConfigBuilder.withClusteringConfig(HoodieClusteringConfig.newBuilder()
.withClusteringPartitionFilterBeginPartition("20211222")
.withClusteringPartitionFilterEndPartition("20211223")
.withClusteringPlanPartitionFilterMode(ClusteringPlanPartitionFilterMode.SELECTED_PARTITIONS)
.build())
.build();
PartitionAwareClusteringPlanStrategy sg = new SparkSizeBasedClusteringPlanStrategy(table, context, config);
ArrayList<String> fakeTimeBasedPartitionsPath = new ArrayList<>();
fakeTimeBasedPartitionsPath.add("20211220");
fakeTimeBasedPartitionsPath.add("20211221");
fakeTimeBasedPartitionsPath.add("20211222");
fakeTimeBasedPartitionsPath.add("20211224");
List list = sg.filterPartitionPaths(fakeTimeBasedPartitionsPath);
assertEquals(1, list.size());
assertSame("20211222", list.get(0));
}
}