1
0

[HUDI-427] [HUDI-971] Implement CLI support for performing bootstrap (#1869)

* [HUDI-971] Clean partitions & fileIds returned by HFileBootstrapIndex
* [HUDI-427] Implement CLI support for performing bootstrap

Co-authored-by: Wenning Ding <wenningd@amazon.com>
Co-authored-by: Balaji Varadarajan <vbalaji@apache.org>
This commit is contained in:
wenningd
2020-08-08 12:37:29 -07:00
committed by GitHub
parent 5ee676e34f
commit 9fe2d2b14a
11 changed files with 448 additions and 28 deletions

View File

@@ -193,7 +193,7 @@ public class BootstrapCommitActionExecutor<T extends HoodieRecordPayload<T>>
}
/**
* Perform Metadata Bootstrap.
* Perform Full Bootstrap.
* @param partitionFilesList List of partitions and files within that partitions
*/
protected Option<HoodieWriteMetadata> fullBootstrap(List<Pair<String, List<HoodieFileStatus>>> partitionFilesList) {

View File

@@ -522,8 +522,10 @@ public class TestHoodieMergeOnReadTable extends HoodieClientTestHarness {
secondClient.rollback(commitTime1);
allFiles = listAllDataFilesInPath(hoodieTable, cfg.getBasePath());
// After rollback, there should be no base file with the failed commit time
assertEquals(0, Arrays.stream(allFiles)
.filter(file -> file.getPath().getName().contains(commitTime1)).count());
List<String> remainingFiles = Arrays.stream(allFiles).filter(file -> file.getPath().getName()
.contains(commitTime1)).map(fileStatus -> fileStatus.getPath().toString()).collect(Collectors.toList());
assertEquals(0, remainingFiles.size(), "There files should have been rolled-back "
+ "when rolling back commit " + commitTime1 + " but are still remaining. Files: " + remainingFiles);
dataFiles = tableView.getLatestBaseFiles().map(HoodieBaseFile::getPath).collect(Collectors.toList());
recordsRead = HoodieMergeOnReadTestUtils.getRecordsUsingInputFormat(hadoopConf, dataFiles, basePath);
assertEquals(200, recordsRead.size());