lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-3758] Fix duplicate fileId error in MOR table type with flink bucket hash Index (#5185)

Browse Source 
* fix duplicate fileId with bucket Index
* replace to load FileGroup from FileSystemView
This commit is contained in:
吴祥平
2022-04-29 14:10:20 +08:00
committed by GitHub
parent b27e8b51d8
commit e421d536ea
1 changed files with 3 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
hudi-flink-datasource/hudi-flink/src/main/java/org/apache/hudi/sink/bucket/BucketStreamWriteFunction.java
View File

@@ -75,11 +75,6 @@ public class BucketStreamWriteFunction<I> extends StreamWriteFunction<I> {
*/
private Set<String> incBucketIndex;
/**
* Returns whether this is an empty table.
*/
private boolean isEmptyTable;
/**
* Constructs a BucketStreamWriteFunction.
*
@@ -99,7 +94,6 @@ public class BucketStreamWriteFunction<I> extends StreamWriteFunction<I> {
this.bucketToLoad = getBucketToLoad();
this.bucketIndex = new HashMap<>();
this.incBucketIndex = new HashSet<>();
this.isEmptyTable = !this.metaClient.getActiveTimeline().filterCompletedInstants().lastInstant().isPresent();
}
@Override
@@ -162,7 +156,7 @@ public class BucketStreamWriteFunction<I> extends StreamWriteFunction<I> {
* This is a required operation for each restart to avoid having duplicate file ids for one bucket.
*/
private void bootstrapIndexIfNeed(String partition) {
if (isEmptyTable || bucketIndex.containsKey(partition)) {
if (bucketIndex.containsKey(partition)) {
return;
}
LOG.info(String.format("Loading Hoodie Table %s, with path %s", this.metaClient.getTableConfig().getTableName(),
@@ -170,8 +164,8 @@ public class BucketStreamWriteFunction<I> extends StreamWriteFunction<I> {
// Load existing fileID belongs to this task
Map<Integer, String> bucketToFileIDMap = new HashMap<>();
this.writeClient.getHoodieTable().getHoodieView().getLatestFileSlices(partition).forEach(fileSlice -> {
String fileID = fileSlice.getFileId();
this.writeClient.getHoodieTable().getFileSystemView().getAllFileGroups(partition).forEach(fileGroup -> {
String fileID = fileGroup.getFileGroupId().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));