lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

[HUDI-3461] The archived timeline for flink streaming reader should not be reused (#4861)

Browse Source 
* Before the patch, the flink streaming reader caches the meta client thus the archived timeline,
  when fetching the instant details from the reused timeline, the exception throws
* Add a method in HoodieTableMetaClient to return a fresh new archived timeline each time
This commit is contained in:
Danny Chan
2022-02-22 15:54:29 +08:00
committed by GitHub
parent 4d1f74ebea
commit 4affdd0c8f
3 changed files with 49 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
hudi-common/src/main/java/org/apache/hudi/common/table/HoodieTableMetaClient.java
View File

@@ -156,7 +156,7 @@ public class HoodieTableMetaClient implements Serializable {
*/
private void readObject(java.io.ObjectInputStream in) throws IOException, ClassNotFoundException {
in.defaultReadObject();
fs = null; // will be lazily inited
fs = null; // will be lazily initialized
}
private void writeObject(java.io.ObjectOutputStream out) throws IOException {
@@ -330,7 +330,7 @@ public class HoodieTableMetaClient implements Serializable {
* Get the archived commits as a timeline. This is costly operation, as all data from the archived files are read.
* This should not be used, unless for historical debugging purposes.
*
* @return Active commit timeline
* @return Archived commit timeline
*/
public synchronized HoodieArchivedTimeline getArchivedTimeline() {
if (archivedTimeline == null) {
@@ -339,6 +339,20 @@ public class HoodieTableMetaClient implements Serializable {
return archivedTimeline;
}
/**
* Returns fresh new archived commits as a timeline from startTs (inclusive).
*
* <p>This is costly operation if really early endTs is specified.
* Be caution to use this only when the time range is short.
*
* <p>This method is not thread safe.
*
* @return Archived commit timeline
*/
public HoodieArchivedTimeline getArchivedTimeline(String startTs) {
return new HoodieArchivedTimeline(this, startTs);
}
/**
* Validate table properties.
* @param properties Properties from writeConfig.

34
hudi-common/src/main/java/org/apache/hudi/common/table/timeline/HoodieArchivedTimeline.java
View File

@@ -79,13 +79,13 @@ public class HoodieArchivedTimeline extends HoodieDefaultTimeline {
private static final String ACTION_TYPE_KEY = "actionType";
private static final String ACTION_STATE = "actionState";
private HoodieTableMetaClient metaClient;
private Map<String, byte[]> readCommits = new HashMap<>();
private final Map<String, byte[]> readCommits = new HashMap<>();
private static final Logger LOG = LogManager.getLogger(HoodieArchivedTimeline.class);
/**
* Loads instants between (startTs, endTs].
* Note that there is no lazy loading, so this may not work if really long time range (endTs-startTs) is specified.
* Loads all the archived instants.
* Note that there is no lazy loading, so this may not work if the archived timeline range is really long.
* TBD: Should we enforce maximum time range?
*/
public HoodieArchivedTimeline(HoodieTableMetaClient metaClient) {
@@ -96,6 +96,19 @@ public class HoodieArchivedTimeline extends HoodieDefaultTimeline {
this.details = (Function<HoodieInstant, Option<byte[]>> & Serializable) this::getInstantDetails;
}
/**
* Loads completed instants from startTs(inclusive).
* Note that there is no lazy loading, so this may not work if really early startTs is specified.
*/
public HoodieArchivedTimeline(HoodieTableMetaClient metaClient, String startTs) {
this.metaClient = metaClient;
setInstants(loadInstants(new StartTsFilter(startTs), true,
record -> HoodieInstant.State.COMPLETED.toString().equals(record.get(ACTION_STATE).toString())));
// multiple casts will make this lambda serializable -
// http://docs.oracle.com/javase/specs/jls/se8/html/jls-15.html#jls-15.16
this.details = (Function<HoodieInstant, Option<byte[]>> & Serializable) this::getInstantDetails;
}
/**
* For serialization and de-serialization only.
*
@@ -300,6 +313,19 @@ public class HoodieArchivedTimeline extends HoodieDefaultTimeline {
}
}
private static class StartTsFilter extends TimeRangeFilter {
private final String startTs;
public StartTsFilter(String startTs) {
super(startTs, null); // endTs is never used
this.startTs = startTs;
}
public boolean isInRange(HoodieInstant instant) {
return HoodieTimeline.compareTimestamps(instant.getTimestamp(), GREATER_THAN_OR_EQUALS, startTs);
}
}
/**
* Sort files by reverse order of version suffix in file name.
*/
@@ -330,7 +356,7 @@ public class HoodieArchivedTimeline extends HoodieDefaultTimeline {
// filter in-memory instants
Set<String> validActions = CollectionUtils.createSet(COMMIT_ACTION, DELTA_COMMIT_ACTION, COMPACTION_ACTION, REPLACE_COMMIT_ACTION);
return new HoodieDefaultTimeline(getInstants().filter(i ->
readCommits.keySet().contains(i.getTimestamp()))
readCommits.containsKey(i.getTimestamp()))
.filter(s -> validActions.contains(s.getAction())), details);
}
}