lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-1529] Add block size to the FileStatus objects returned from metadata table to avoid too many file splits (#2451)

Browse Source
This commit is contained in:
Udit Mehrotra
2021-01-18 07:29:53 -08:00
committed by GitHub
parent 3d1d5d00b0
commit 684e12e9fc
3 changed files with 16 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
hudi-common/src/main/java/org/apache/hudi/metadata/BaseTableMetadata.java
View File

@@ -202,7 +202,7 @@ public abstract class BaseTableMetadata implements HoodieTableMetadata {
throw new HoodieMetadataException("Metadata record for partition " + partitionName + " is inconsistent: "
+ hoodieRecord.get().getData());
}
statuses = hoodieRecord.get().getData().getFileStatuses(partitionPath);
statuses = hoodieRecord.get().getData().getFileStatuses(hadoopConf.get(), partitionPath);
}
if (validateLookups) {

10
hudi-common/src/main/java/org/apache/hudi/metadata/HoodieMetadataPayload.java
View File

@@ -29,7 +29,9 @@ import org.apache.hudi.exception.HoodieMetadataException;
import org.apache.avro.Schema;
import org.apache.avro.generic.GenericRecord;
import org.apache.avro.generic.IndexedRecord;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.FileStatus;
import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import java.io.IOException;
@@ -177,10 +179,12 @@ public class HoodieMetadataPayload implements HoodieRecordPayload<HoodieMetadata
/**
* Returns the files added as part of this record.
*/
public FileStatus[] getFileStatuses(Path partitionPath) {
public FileStatus[] getFileStatuses(Configuration hadoopConf, Path partitionPath) throws IOException {
FileSystem fs = partitionPath.getFileSystem(hadoopConf);
long blockSize = fs.getDefaultBlockSize(partitionPath);
return filterFileInfoEntries(false)
.map(e -> new FileStatus(e.getValue().getSize(), false, 0, 0, 0, 0, null, null, null,
new Path(partitionPath, e.getKey())))
.map(e -> new FileStatus(e.getValue().getSize(), false, 0, blockSize, 0, 0,
null, null, null, new Path(partitionPath, e.getKey())))
.toArray(FileStatus[]::new);
}