lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

Table rollback for inflight compactions MUST not delete instant files at any time to avoid race conditions

Browse Source
This commit is contained in:
Balaji Varadarajan
2019-02-05 16:54:31 -08:00
committed by vinoth chandar
parent defcf6a0b9
commit 8adaca3454
6 changed files with 97 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
hoodie-client/src/main/java/com/uber/hoodie/HoodieWriteClient.java
View File

@@ -839,7 +839,7 @@ public class HoodieWriteClient<T extends HoodieRecordPayload> implements Seriali
// Remove interleaving pending compactions before rolling back commits
pendingCompactionToRollback.forEach(this::deletePendingCompaction);
List<HoodieRollbackStat> stats = table.rollback(jsc, commitsToRollback);
List<HoodieRollbackStat> stats = table.rollback(jsc, commitsToRollback, true);
// cleanup index entries
commitsToRollback.forEach(s -> {
@@ -1206,8 +1206,9 @@ public class HoodieWriteClient<T extends HoodieRecordPayload> implements Seriali
* @param inflightInstant Inflight Compaction Instant
* @param table Hoodie Table
*/
private void rollbackInflightCompaction(HoodieInstant inflightInstant, HoodieTable table) throws IOException {
table.rollback(jsc, ImmutableList.copyOf(new String[] { inflightInstant.getTimestamp() }));
@VisibleForTesting
void rollbackInflightCompaction(HoodieInstant inflightInstant, HoodieTable table) throws IOException {
table.rollback(jsc, ImmutableList.copyOf(new String[] { inflightInstant.getTimestamp() }), false);
// Revert instant state file
table.getActiveTimeline().revertCompactionInflightToRequested(inflightInstant);
}

26
hoodie-client/src/main/java/com/uber/hoodie/table/HoodieCopyOnWriteTable.java
View File

@@ -347,7 +347,7 @@ public class HoodieCopyOnWriteTable<T extends HoodieRecordPayload> extends Hoodi
}
@Override
public List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits)
public List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits, boolean deleteInstants)
throws IOException {
String actionType = metaClient.getCommitActionType();
HoodieActiveTimeline activeTimeline = this.getActiveTimeline();
@@ -375,13 +375,29 @@ public class HoodieCopyOnWriteTable<T extends HoodieRecordPayload> extends Hoodi
// clean temporary data files
cleanTemporaryDataFiles(jsc);
// Remove the rolled back inflight commits
commits.stream().map(s -> new HoodieInstant(true, actionType, s))
.forEach(activeTimeline::deleteInflight);
logger.info("Deleted inflight commits " + commits);
// Delete Inflight instants if enabled
deleteInflightInstants(deleteInstants, activeTimeline,
commits.stream().map(s -> new HoodieInstant(true, actionType, s)).collect(Collectors.toList()));
return stats;
}
/**
* Delete Inflight instants if enabled
* @param deleteInstants Enable Deletion of Inflight instants
* @param activeTimeline Hoodie active timeline
* @param instantsToBeDeleted Instants to be deleted
*/
protected static void deleteInflightInstants(boolean deleteInstants, HoodieActiveTimeline activeTimeline,
List<HoodieInstant> instantsToBeDeleted) {
// Remove the rolled back inflight commits
if (deleteInstants) {
instantsToBeDeleted.forEach(activeTimeline::deleteInflight);
logger.info("Deleted inflight commits " + instantsToBeDeleted);
} else {
logger.warn("Rollback finished without deleting inflight instant files. Instants=" + instantsToBeDeleted);
}
}
/**
* Finalize the written data files
*

14
hoodie-client/src/main/java/com/uber/hoodie/table/HoodieMergeOnReadTable.java
View File

@@ -165,7 +165,7 @@ public class HoodieMergeOnReadTable<T extends HoodieRecordPayload> extends
}
@Override
public List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits)
public List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits, boolean deleteInstants)
throws IOException {
//At the moment, MOR table type does not support nested rollbacks
@@ -274,11 +274,13 @@ public class HoodieMergeOnReadTable<T extends HoodieRecordPayload> extends
return hoodieRollbackStats;
}).collect(Collectors.toList())).flatMap(List::iterator).filter(Objects::nonNull).collect();
commitsAndCompactions.entrySet().stream().map(
entry -> new HoodieInstant(true, entry.getValue().getAction(),
entry.getValue().getTimestamp())).forEach(this.getActiveTimeline()::deleteInflight);
logger
.debug("Time(in ms) taken to finish rollback " + (System.currentTimeMillis() - startTime));
// Delete Inflight instants if enabled
deleteInflightInstants(deleteInstants, this.getActiveTimeline(),
commitsAndCompactions.entrySet().stream().map(
entry -> new HoodieInstant(true, entry.getValue().getAction(), entry.getValue().getTimestamp()))
.collect(Collectors.toList()));
logger.debug("Time(in ms) taken to finish rollback " + (System.currentTimeMillis() - startTime));
return allRollbackStats;
}

4
hoodie-client/src/main/java/com/uber/hoodie/table/HoodieTable.java
View File

@@ -251,9 +251,9 @@ public abstract class HoodieTable<T extends HoodieRecordPayload> implements Seri
/**
* Rollback the (inflight/committed) record changes with the given commit time. Four steps: (1)
* Atomically unpublish this commit (2) clean indexing data (3) clean new generated parquet files
* / log blocks (4) Finally, delete .<action>.commit or .<action>.inflight file
* / log blocks (4) Finally, delete .<action>.commit or .<action>.inflight file if deleteInstants = true
*/
public abstract List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits)
public abstract List<HoodieRollbackStat> rollback(JavaSparkContext jsc, List<String> commits, boolean deleteInstants)
throws IOException;
/**