1
0

[HUDI-1999] Refresh the base file view cache for WriteProfile (#3067)

Refresh the view to discover new small files.
This commit is contained in:
Danny Chan
2021-06-15 23:18:38 +08:00
committed by GitHub
parent f922837064
commit cb642ceb75
7 changed files with 197 additions and 61 deletions

View File

@@ -55,6 +55,11 @@ public class FlinkMergeAndReplaceHandle<T extends HoodieRecordPayload, I, K, O>
private boolean isClosed = false;
/**
* Flag saying whether we should replace the old file with new.
*/
private boolean shouldReplace = true;
public FlinkMergeAndReplaceHandle(HoodieWriteConfig config, String instantTime, HoodieTable<T, I, K, O> hoodieTable,
Iterator<HoodieRecord<T>> recordItr, String partitionPath, String fileId,
TaskContextSupplier taskContextSupplier, Path basePath) {
@@ -103,11 +108,12 @@ public class FlinkMergeAndReplaceHandle<T extends HoodieRecordPayload, I, K, O>
@Override
protected void makeOldAndNewFilePaths(String partitionPath, String oldFileName, String newFileName) {
// old and new file name expects to be the same.
if (!oldFileName.equals(newFileName)) {
if (!FSUtils.getCommitTime(oldFileName).equals(instantTime)) {
LOG.warn("MERGE and REPLACE handle expect the same name for old and new files,\n"
+ "while got new file: " + newFileName + " with old file: " + oldFileName + ",\n"
+ "this rarely happens when the checkpoint success event was not received yet\n"
+ "but the write task flush with new instant time, which does not break the UPSERT semantics");
shouldReplace = false;
}
super.makeOldAndNewFilePaths(partitionPath, oldFileName, newFileName);
try {
@@ -146,6 +152,10 @@ public class FlinkMergeAndReplaceHandle<T extends HoodieRecordPayload, I, K, O>
}
public void finalizeWrite() {
// Behaves like the normal merge handle if the write instant time changes.
if (!shouldReplace) {
return;
}
// The file visibility should be kept by the configured ConsistencyGuard instance.
try {
fs.delete(oldFilePath, false);

View File

@@ -206,11 +206,11 @@ public class StreamWriteOperatorCoordinator
}
// start new instant.
startInstant();
// sync Hive if is enabled
syncHiveIfEnabled();
}
}, "commits the instant %s", this.instant
);
// sync Hive if is enabled
syncHiveIfEnabled();
}
private void syncHiveIfEnabled() {

View File

@@ -54,12 +54,13 @@ public class DeltaWriteProfile extends WriteProfile {
// Find out all eligible small file slices
if (!commitTimeline.empty()) {
HoodieInstant latestCommitTime = commitTimeline.lastInstant().get();
// initialize the filesystem view based on the commit metadata
initFSViewIfNecessary(commitTimeline);
// find smallest file in partition and append to it
List<FileSlice> allSmallFileSlices = new ArrayList<>();
// If we can index log files, we can add more inserts to log files for fileIds including those under
// pending compaction.
List<FileSlice> allFileSlices =
table.getSliceView().getLatestFileSlicesBeforeOrOn(partitionPath, latestCommitTime.getTimestamp(), true)
List<FileSlice> allFileSlices = fsView.getLatestFileSlicesBeforeOrOn(partitionPath, latestCommitTime.getTimestamp(), true)
.collect(Collectors.toList());
for (FileSlice fileSlice : allFileSlices) {
if (isSmallFile(fileSlice)) {

View File

@@ -25,12 +25,17 @@ import org.apache.hudi.common.model.HoodieCommitMetadata;
import org.apache.hudi.common.model.HoodieRecordLocation;
import org.apache.hudi.common.table.timeline.HoodieInstant;
import org.apache.hudi.common.table.timeline.HoodieTimeline;
import org.apache.hudi.common.table.view.HoodieTableFileSystemView;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.sink.partitioner.BucketAssigner;
import org.apache.hudi.table.HoodieFlinkTable;
import org.apache.hudi.table.HoodieTable;
import org.apache.hudi.table.action.commit.SmallFile;
import org.apache.hudi.util.StreamerUtil;
import org.apache.flink.core.fs.Path;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileStatus;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
@@ -55,6 +60,11 @@ public class WriteProfile {
*/
protected final HoodieWriteConfig config;
/**
* Table base path.
*/
private final Path basePath;
/**
* The hoodie table.
*/
@@ -81,11 +91,23 @@ public class WriteProfile {
*/
private long reloadedCheckpointId;
/**
* The file system view cache for one checkpoint interval.
*/
protected HoodieTableFileSystemView fsView;
/**
* Hadoop configuration.
*/
private final Configuration hadoopConf;
public WriteProfile(HoodieWriteConfig config, HoodieFlinkEngineContext context) {
this.config = config;
this.basePath = new Path(config.getBasePath());
this.smallFilesMap = new HashMap<>();
this.recordsPerBucket = config.getCopyOnWriteInsertSplitSize();
this.table = HoodieFlinkTable.create(config, context);
this.hadoopConf = StreamerUtil.getHadoopConf();
// profile the record statistics on construction
recordProfile();
}
@@ -160,7 +182,9 @@ public class WriteProfile {
if (!commitTimeline.empty()) { // if we have some commits
HoodieInstant latestCommitTime = commitTimeline.lastInstant().get();
List<HoodieBaseFile> allFiles = table.getBaseFileOnlyView()
// initialize the filesystem view based on the commit metadata
initFSViewIfNecessary(commitTimeline);
List<HoodieBaseFile> allFiles = fsView
.getLatestBaseFilesBeforeOrOn(partitionPath, latestCommitTime.getTimestamp()).collect(Collectors.toList());
for (HoodieBaseFile file : allFiles) {
@@ -178,6 +202,16 @@ public class WriteProfile {
return smallFileLocations;
}
protected void initFSViewIfNecessary(HoodieTimeline commitTimeline) {
if (fsView == null) {
List<HoodieCommitMetadata> metadataList = commitTimeline.getInstants()
.map(instant -> WriteProfiles.getCommitMetadata(config.getTableName(), basePath, instant, commitTimeline))
.collect(Collectors.toList());
FileStatus[] commitFiles = WriteProfiles.getWritePathsOfInstants(basePath, hadoopConf, metadataList);
fsView = new HoodieTableFileSystemView(table.getMetaClient(), commitTimeline, commitFiles);
}
}
private void recordProfile() {
this.avgSize = averageBytesPerRecord();
if (config.shouldAllowMultiWriteOnSameInstant()) {
@@ -200,6 +234,7 @@ public class WriteProfile {
return;
}
recordProfile();
this.fsView = null;
this.smallFilesMap.clear();
this.table.getMetaClient().reloadActiveTimeline();
this.reloadedCheckpointId = checkpointId;

View File

@@ -19,15 +19,33 @@
package org.apache.hudi.sink.partitioner.profile;
import org.apache.hudi.client.common.HoodieFlinkEngineContext;
import org.apache.hudi.common.fs.FSUtils;
import org.apache.hudi.common.model.HoodieCommitMetadata;
import org.apache.hudi.common.table.timeline.HoodieInstant;
import org.apache.hudi.common.table.timeline.HoodieTimeline;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.exception.HoodieException;
import org.apache.flink.core.fs.Path;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.io.IOException;
import java.util.Collection;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
/**
* Factory for {@link WriteProfile}.
*/
public class WriteProfiles {
private static final Logger LOG = LoggerFactory.getLogger(WriteProfiles.class);
private static final Map<String, WriteProfile> PROFILES = new HashMap<>();
private WriteProfiles() {}
@@ -58,4 +76,80 @@ public class WriteProfiles {
public static void clean(String path) {
PROFILES.remove(path);
}
/**
* Returns all the incremental write file path statuses with the given commits metadata.
*
* @param basePath Table base path
* @param hadoopConf The hadoop conf
* @param metadataList The commits metadata
* @return the file statuses array
*/
public static FileStatus[] getWritePathsOfInstants(
Path basePath,
Configuration hadoopConf,
List<HoodieCommitMetadata> metadataList) {
FileSystem fs = FSUtils.getFs(basePath.toString(), hadoopConf);
return metadataList.stream().map(metadata -> getWritePathsOfInstant(basePath, metadata, fs))
.flatMap(Collection::stream).toArray(FileStatus[]::new);
}
/**
* Returns the commit file paths with given metadata.
*
* @param basePath Table base path
* @param metadata The metadata
* @param fs The filesystem
*
* @return the commit file status list
*/
private static List<FileStatus> getWritePathsOfInstant(Path basePath, HoodieCommitMetadata metadata, FileSystem fs) {
return metadata.getFileIdAndFullPaths(basePath.toString()).values().stream()
.map(org.apache.hadoop.fs.Path::new)
// filter out the file paths that does not exist, some files may be cleaned by
// the cleaner.
.filter(path -> {
try {
return fs.exists(path);
} catch (IOException e) {
LOG.error("Checking exists of path: {} error", path);
throw new HoodieException(e);
}
}).map(path -> {
try {
return fs.getFileStatus(path);
} catch (IOException e) {
LOG.error("Get write status of path: {} error", path);
throw new HoodieException(e);
}
})
// filter out crushed files
.filter(fileStatus -> fileStatus.getLen() > 0)
.collect(Collectors.toList());
}
/**
* Returns the commit metadata of the given instant.
*
* @param tableName The table name
* @param basePath The table base path
* @param instant The hoodie instant
* @param timeline The timeline
*
* @return the commit metadata
*/
public static HoodieCommitMetadata getCommitMetadata(
String tableName,
Path basePath,
HoodieInstant instant,
HoodieTimeline timeline) {
byte[] data = timeline.getInstantDetails(instant).get();
try {
return HoodieCommitMetadata.fromBytes(data, HoodieCommitMetadata.class);
} catch (IOException e) {
LOG.error("Get write metadata for table {} with instant {} and path: {} error",
tableName, instant.getTimestamp(), basePath);
throw new HoodieException(e);
}
}
}

View File

@@ -18,7 +18,6 @@
package org.apache.hudi.source;
import org.apache.hudi.common.fs.FSUtils;
import org.apache.hudi.common.model.BaseFile;
import org.apache.hudi.common.model.HoodieCommitMetadata;
import org.apache.hudi.common.model.HoodieLogFile;
@@ -30,7 +29,7 @@ import org.apache.hudi.common.table.view.HoodieTableFileSystemView;
import org.apache.hudi.common.util.Option;
import org.apache.hudi.common.util.ValidationUtils;
import org.apache.hudi.configuration.FlinkOptions;
import org.apache.hudi.exception.HoodieException;
import org.apache.hudi.sink.partitioner.profile.WriteProfiles;
import org.apache.hudi.table.format.mor.MergeOnReadInputSplit;
import org.apache.hudi.util.StreamerUtil;
@@ -46,11 +45,9 @@ import org.apache.flink.streaming.api.checkpoint.CheckpointedFunction;
import org.apache.flink.streaming.api.functions.source.RichSourceFunction;
import org.apache.flink.streaming.api.functions.source.SourceFunction;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.io.IOException;
import java.util.ArrayList;
import java.util.Collection;
import java.util.List;
@@ -220,10 +217,11 @@ public class StreamReadMonitoringFunction
// 3. filter the full file paths
// 4. use the file paths from #step 3 as the back-up of the filesystem view
String tableName = conf.getString(FlinkOptions.TABLE_NAME);
List<HoodieCommitMetadata> metadataList = instants.stream()
.map(instant -> getCommitMetadata(instant, commitTimeline)).collect(Collectors.toList());
.map(instant -> WriteProfiles.getCommitMetadata(tableName, path, instant, commitTimeline)).collect(Collectors.toList());
Set<String> writePartitions = getWritePartitionPaths(metadataList);
FileStatus[] fileStatuses = getWritePathsOfInstants(metadataList);
FileStatus[] fileStatuses = WriteProfiles.getWritePathsOfInstants(path, hadoopConf, metadataList);
if (fileStatuses.length == 0) {
LOG.warn("No files found for reading in user provided path.");
return;
@@ -334,52 +332,4 @@ public class StreamReadMonitoringFunction
.flatMap(Collection::stream)
.collect(Collectors.toSet());
}
/**
* Returns all the incremental write file path statuses with the given commits metadata.
*
* @param metadataList The commits metadata
* @return the file statuses array
*/
private FileStatus[] getWritePathsOfInstants(List<HoodieCommitMetadata> metadataList) {
FileSystem fs = FSUtils.getFs(path.toString(), hadoopConf);
return metadataList.stream().map(metadata -> getWritePathsOfInstant(metadata, fs))
.flatMap(Collection::stream).toArray(FileStatus[]::new);
}
private List<FileStatus> getWritePathsOfInstant(HoodieCommitMetadata metadata, FileSystem fs) {
return metadata.getFileIdAndFullPaths(path.toString()).values().stream()
.map(org.apache.hadoop.fs.Path::new)
// filter out the file paths that does not exist, some files may be cleaned by
// the cleaner.
.filter(path -> {
try {
return fs.exists(path);
} catch (IOException e) {
LOG.error("Checking exists of path: {} error", path);
throw new HoodieException(e);
}
}).map(path -> {
try {
return fs.getFileStatus(path);
} catch (IOException e) {
LOG.error("Get write status of path: {} error", path);
throw new HoodieException(e);
}
})
// filter out crushed files
.filter(fileStatus -> fileStatus.getLen() > 0)
.collect(Collectors.toList());
}
private HoodieCommitMetadata getCommitMetadata(HoodieInstant instant, HoodieTimeline timeline) {
byte[] data = timeline.getInstantDetails(instant).get();
try {
return HoodieCommitMetadata.fromBytes(data, HoodieCommitMetadata.class);
} catch (IOException e) {
LOG.error("Get write metadata for table {} with instant {} and path: {} error",
conf.getString(FlinkOptions.TABLE_NAME), instant.getTimestamp(), path);
throw new HoodieException(e);
}
}
}

View File

@@ -22,6 +22,8 @@ import org.apache.hudi.client.FlinkTaskContextSupplier;
import org.apache.hudi.client.common.HoodieFlinkEngineContext;
import org.apache.hudi.common.config.SerializableConfiguration;
import org.apache.hudi.common.model.HoodieRecordLocation;
import org.apache.hudi.common.table.timeline.HoodieInstant;
import org.apache.hudi.common.util.Option;
import org.apache.hudi.config.HoodieWriteConfig;
import org.apache.hudi.sink.partitioner.profile.WriteProfile;
import org.apache.hudi.table.action.commit.BucketInfo;
@@ -29,6 +31,7 @@ import org.apache.hudi.table.action.commit.BucketType;
import org.apache.hudi.table.action.commit.SmallFile;
import org.apache.hudi.util.StreamerUtil;
import org.apache.hudi.utils.TestConfigurations;
import org.apache.hudi.utils.TestData;
import org.apache.flink.configuration.Configuration;
import org.junit.jupiter.api.BeforeEach;
@@ -45,6 +48,10 @@ import java.util.Map;
import static org.hamcrest.CoreMatchers.is;
import static org.hamcrest.MatcherAssert.assertThat;
import static org.junit.jupiter.api.Assertions.assertFalse;
import static org.junit.jupiter.api.Assertions.assertNotEquals;
import static org.junit.jupiter.api.Assertions.assertNotNull;
import static org.junit.jupiter.api.Assertions.assertTrue;
/**
* Test cases for {@link BucketAssigner}.
@@ -52,6 +59,7 @@ import static org.hamcrest.MatcherAssert.assertThat;
public class TestBucketAssigner {
private HoodieWriteConfig writeConfig;
private HoodieFlinkEngineContext context;
private Configuration conf;
@TempDir
File tempFile;
@@ -59,7 +67,7 @@ public class TestBucketAssigner {
@BeforeEach
public void before() throws IOException {
final String basePath = tempFile.getAbsolutePath();
final Configuration conf = TestConfigurations.getDefaultConf(basePath);
conf = TestConfigurations.getDefaultConf(basePath);
writeConfig = StreamerUtil.getHoodieClientConfig(conf);
context = new HoodieFlinkEngineContext(
@@ -291,6 +299,44 @@ public class TestBucketAssigner {
assertBucketEquals(bucketInfo2, "par1", BucketType.UPDATE, "f1");
}
@Test
public void testWriteProfileReload() throws Exception {
WriteProfile writeProfile = new WriteProfile(writeConfig, context);
List<SmallFile> smallFiles1 = writeProfile.getSmallFiles("par1");
assertTrue(smallFiles1.isEmpty(), "Should have no small files");
TestData.writeData(TestData.DATA_SET_INSERT, conf);
Option<String> instantOption = getLastCompleteInstant(writeProfile);
assertFalse(instantOption.isPresent());
writeProfile.reload(1);
String instant1 = getLastCompleteInstant(writeProfile).orElse(null);
assertNotNull(instant1);
List<SmallFile> smallFiles2 = writeProfile.getSmallFiles("par1");
assertThat("Should have 1 small file", smallFiles2.size(), is(1));
assertThat("Small file should have same timestamp as last complete instant",
smallFiles2.get(0).location.getInstantTime(), is(instant1));
TestData.writeData(TestData.DATA_SET_INSERT, conf);
List<SmallFile> smallFiles3 = writeProfile.getSmallFiles("par1");
assertThat("Should have 1 small file", smallFiles3.size(), is(1));
assertThat("Non-reloaded write profile has the same base file view as before",
smallFiles3.get(0).location.getInstantTime(), is(instant1));
writeProfile.reload(2);
String instant2 = getLastCompleteInstant(writeProfile).orElse(null);
assertNotEquals(instant2, instant1, "Should have new complete instant");
List<SmallFile> smallFiles4 = writeProfile.getSmallFiles("par1");
assertThat("Should have 1 small file", smallFiles4.size(), is(1));
assertThat("Small file should have same timestamp as last complete instant",
smallFiles4.get(0).location.getInstantTime(), is(instant2));
}
private static Option<String> getLastCompleteInstant(WriteProfile profile) {
return profile.getTable().getMetaClient().getCommitsTimeline()
.filterCompletedInstants().lastInstant().map(HoodieInstant::getTimestamp);
}
private void assertBucketEquals(
BucketInfo bucketInfo,
String partition,