1
0

Adding a new Partition/Time based compaction strategy

This commit is contained in:
Nishith Agarwal
2017-10-25 13:30:15 -07:00
committed by vinoth chandar
parent 051f600b7f
commit 4aed5c7338
4 changed files with 98 additions and 6 deletions

View File

@@ -128,7 +128,7 @@ public class TestHoodieCompactor {
compactor.compact(jsc, getConfig(), table, HoodieActiveTimeline.createNewCommitTime());
String basePath = table.getMetaClient().getBasePath();
assertTrue("If there is nothing to compact, result will be empty",
result.getFileIdAndFullPaths(basePath).isEmpty());
result == null || result.getFileIdAndFullPaths(basePath).isEmpty());
}
@Test

View File

@@ -16,25 +16,29 @@
package com.uber.hoodie.io.strategy;
import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertTrue;
import com.beust.jcommander.internal.Lists;
import com.google.common.collect.Maps;
import com.uber.hoodie.config.HoodieCompactionConfig;
import com.uber.hoodie.config.HoodieWriteConfig;
import com.uber.hoodie.io.compact.CompactionOperation;
import com.uber.hoodie.io.compact.strategy.BoundedIOCompactionStrategy;
import com.uber.hoodie.io.compact.strategy.DayBasedCompactionStrategy;
import com.uber.hoodie.io.compact.strategy.LogFileSizeBasedCompactionStrategy;
import com.uber.hoodie.io.compact.strategy.UnBoundedCompactionStrategy;
import org.junit.Test;
import java.util.List;
import java.util.Map;
import java.util.Random;
import java.util.stream.Collectors;
import org.junit.Test;
import static org.junit.Assert.assertEquals;
import static org.junit.Assert.assertTrue;
public class TestHoodieCompactionStrategy {
private static final long MB = 1024 * 1024L;
private String [] partitionPaths = {"2017/01/01", "2017/01/02", "2017/01/03"};
@Test
public void testUnBounded() {
@@ -106,11 +110,36 @@ public class TestHoodieCompactionStrategy {
1204, (long) returnedSize);
}
@Test
public void testPartitionAwareCompactionSimple() {
Map<Long, List<Long>> sizesMap = Maps.newHashMap();
sizesMap.put(120 * MB, Lists.newArrayList(60 * MB, 10 * MB, 80 * MB));
sizesMap.put(110 * MB, Lists.newArrayList());
sizesMap.put(100 * MB, Lists.newArrayList(MB));
sizesMap.put(90 * MB, Lists.newArrayList(1024 * MB));
DayBasedCompactionStrategy strategy = new DayBasedCompactionStrategy();
HoodieWriteConfig writeConfig = HoodieWriteConfig.newBuilder().withPath("/tmp")
.withCompactionConfig(
HoodieCompactionConfig.newBuilder().withCompactionStrategy(strategy)
.withTargetIOPerCompactionInMB(400).build()).build();
List<CompactionOperation> operations = createCompactionOperations(writeConfig, sizesMap);
List<CompactionOperation> returned = strategy.orderAndFilter(writeConfig, operations);
assertTrue("DayBasedCompactionStrategy should have resulted in fewer compactions",
returned.size() < operations.size());
int comparision = strategy.getComparator().compare(returned.get(returned.size() - 1), returned.get(0));
// Either the partition paths are sorted in descending order or they are equal
assertTrue("DayBasedCompactionStrategy should sort partitions in descending order",
comparision >= 0);
}
private List<CompactionOperation> createCompactionOperations(HoodieWriteConfig config,
Map<Long, List<Long>> sizesMap) {
List<CompactionOperation> operations = Lists.newArrayList(sizesMap.size());
sizesMap.forEach((k, v) -> {
operations.add(new CompactionOperation(TestHoodieDataFile.newDataFile(k), "",
operations.add(new CompactionOperation(TestHoodieDataFile.newDataFile(k),
partitionPaths[new Random().nextInt(partitionPaths.length - 1)],
v.stream().map(TestHoodieLogFile::newLogFile).collect(
Collectors.toList()), config));
});