Skip to content

[HUDI-3539] Flink bucket index bucketID bootstrap optimization. #5093

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
garyli1019 merged 3 commits into from
Mar 28, 2022
Merged

[HUDI-3539] Flink bucket index bucketID bootstrap optimization. #5093

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
0613c51
[HUDI-3539] Flink bucket index bucketID bootstrap optimization.
Mar 22, 2022
feb745c
fix comments
Mar 22, 2022
7d10408
improvement
Mar 24, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
97 changes: 57 additions & 40 deletions ...ource/hudi-flink/src/main/java/org/apache/hudi/sink/bucket/BucketStreamWriteFunction.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,6 +40,8 @@
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Set;

import static java.util.stream.Collectors.toList;

Expand Down Expand Up @@ -67,16 +69,22 @@ public class BucketStreamWriteFunction<I> extends StreamWriteFunction<I> {
private String indexKeyFields;

/**
* BucketID to file group mapping.
* BucketID should be load in this task.
*/
private HashMap<String, String> bucketIndex;
private Set<Integer> bucketToLoad;

/**
* BucketID to file group mapping in each partition.
* Map(partition -> Map(bucketId, fileID)).
*/
private Map<String, Map<Integer, String>> bucketIndex;

/**
* Incremental bucket index of the current checkpoint interval,
* it is needed because the bucket type('I' or 'U') should be decided based on the committed files view,
* all the records in one bucket should have the same bucket type.
*/
private HashMap<String, String> incBucketIndex;
private Set<String> incBucketIndex;

/**
* Constructs a BucketStreamWriteFunction.
Expand All @@ -95,9 +103,10 @@ public void open(Configuration parameters) throws IOException {
this.taskID = getRuntimeContext().getIndexOfThisSubtask();
this.parallelism = getRuntimeContext().getNumberOfParallelSubtasks();
this.maxParallelism = getRuntimeContext().getMaxNumberOfParallelSubtasks();
this.bucketToLoad = new HashSet<>();
this.bucketIndex = new HashMap<>();
this.incBucketIndex = new HashMap<>();
bootstrapIndex();
this.incBucketIndex = new HashSet<>();
getBucketToLoad();
}

@Override
Expand All @@ -109,25 +118,30 @@ public void initializeState(FunctionInitializationContext context) throws Except
@Override
public void snapshotState() {
super.snapshotState();
this.bucketIndex.putAll(this.incBucketIndex);
this.incBucketIndex.clear();
}

@Override
public void processElement(I i, ProcessFunction<I, Object>.Context context, Collector<Object> collector) throws Exception {
HoodieRecord<?> record = (HoodieRecord<?>) i;
final HoodieKey hoodieKey = record.getKey();
final String partition = hoodieKey.getPartitionPath();
final HoodieRecordLocation location;

bootstrapIndexIfNeed(partition);
Map<Integer, String> bucketToFileIdMap = bucketIndex.get(partition);
final int bucketNum = BucketIdentifier.getBucketId(hoodieKey, indexKeyFields, this.bucketNum);
Copy link
Contributor

@danny0405 danny0405 Mar 28, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

.get(partition) -> computeIfAbsent(partition, p -> new HashMap<>())

final String partitionBucketId = BucketIdentifier.partitionBucketIdStr(hoodieKey.getPartitionPath(), bucketNum);

if (bucketIndex.containsKey(partitionBucketId)) {
location = new HoodieRecordLocation("U", bucketIndex.get(partitionBucketId));
if (incBucketIndex.contains(partitionBucketId)) {
location = new HoodieRecordLocation("I", bucketToFileIdMap.get(bucketNum));
} else if (bucketToFileIdMap.containsKey(bucketNum)) {
location = new HoodieRecordLocation("U", bucketToFileIdMap.get(bucketNum));
} else {
String newFileId = BucketIdentifier.newBucketFileIdPrefix(bucketNum);
location = new HoodieRecordLocation("I", newFileId);
incBucketIndex.put(partitionBucketId, newFileId);
bucketToFileIdMap.put(bucketNum,newFileId);
incBucketIndex.add(partitionBucketId);
}
record.unseal();
record.setCurrentLocation(location);
Expand All @@ -136,17 +150,10 @@ public void processElement(I i, ProcessFunction<I, Object>.Context context, Coll
}

/**
* Get partition_bucket -> fileID mapping from the existing hudi table.
* This is a required operation for each restart to avoid having duplicate file ids for one bucket.
* Bootstrap bucket info from existing file system,
* bucketNum % totalParallelism == this taskID belongs to this task.
*/
private void bootstrapIndex() throws IOException {
Option<HoodieInstant> latestCommitTime = table.getFileSystemView().getTimeline().filterCompletedInstants().lastInstant();
if (!latestCommitTime.isPresent()) {
return;
}
// bootstrap bucket info from existing file system
// bucketNum % totalParallelism == this taskID belongs to this task
HashSet<Integer> bucketToLoad = new HashSet<>();
private void getBucketToLoad() {
for (int i = 0; i < bucketNum; i++) {
int partitionOfBucket = BucketIdentifier.mod(i, parallelism);
if (partitionOfBucket == taskID) {
Expand All @@ -157,31 +164,41 @@ private void bootstrapIndex() throws IOException {
}
}
bucketToLoad.forEach(bucket -> LOG.info(String.format("bucketToLoad contains %s", bucket)));
}

/**
* Get partition_bucket -> fileID mapping from the existing hudi table.
* This is a required operation for each restart to avoid having duplicate file ids for one bucket.
*/
private void bootstrapIndexIfNeed(String partition) throws IOException {
if (bucketIndex.containsKey(partition)) {
return;
}
Option<HoodieInstant> latestCommitTime = table.getHoodieView().getTimeline().filterCompletedInstants().lastInstant();
if (!latestCommitTime.isPresent()) {
bucketIndex.put(partition, new HashMap<>());
return;
}
LOG.info(String.format("Loading Hoodie Table %s, with path %s", table.getMetaClient().getTableConfig().getTableName(),
table.getMetaClient().getBasePath()));

// Iterate through all existing partitions to load existing fileID belongs to this task
List<String> partitions = table.getMetadata().getAllPartitionPaths();
for (String partitionPath : partitions) {
List<FileSlice> latestFileSlices = table.getSliceView()
.getLatestFileSlices(partitionPath)
.collect(toList());
for (FileSlice fileslice : latestFileSlices) {
String fileID = fileslice.getFileId();
int bucketNumber = BucketIdentifier.bucketIdFromFileId(fileID);
if (bucketToLoad.contains(bucketNumber)) {
String partitionBucketId = BucketIdentifier.partitionBucketIdStr(partitionPath, bucketNumber);
LOG.info(String.format("Should load this partition bucket %s with fileID %s", partitionBucketId, fileID));
if (bucketIndex.containsKey(partitionBucketId)) {
throw new RuntimeException(String.format("Duplicate fileID %s from partitionBucket %s found "
+ "during the BucketStreamWriteFunction index bootstrap.", fileID, partitionBucketId));
} else {
LOG.info(String.format("Adding fileID %s to the partition bucket %s.", fileID, partitionBucketId));
bucketIndex.put(partitionBucketId, fileID);
}
table.getMetaClient().getBasePath() + "/" + partition));

// Load existing fileID belongs to this task
Map<Integer, String> bucketToFileIDMap = new HashMap<>();
List<FileSlice> fileSlices = table.getHoodieView().getLatestFileSlices(partition).collect(toList());
for (FileSlice fileSlice : fileSlices) {
String fileID = fileSlice.getFileId();
int bucketNumber = BucketIdentifier.bucketIdFromFileId(fileID);
if (bucketToLoad.contains(bucketNumber)) {
LOG.info(String.format("Should load this partition bucket %s with fileID %s", bucketNumber, fileID));
if (bucketToFileIDMap.containsKey(bucketNumber)) {
throw new RuntimeException(String.format("Duplicate fileID %s from bucket %s of partition %s found "
+ "during the BucketStreamWriteFunction index bootstrap.", fileID, bucketNumber, partition));
} else {
LOG.info(String.format("Adding fileID %s to the bucket %s of partition %s.", fileID, bucketNumber, partition));
bucketToFileIDMap.put(bucketNumber, fileID);
}
}
}
bucketIndex.put(partition, bucketToFileIDMap);
}
}