SteNicholas commented on a change in pull request #3779:
URL: https://github.com/apache/hudi/pull/3779#discussion_r727103868
##########
File path:
hudi-client/hudi-flink-client/src/main/java/org/apache/hudi/client/HoodieFlinkWriteClient.java
##########
@@ -257,6 +260,11 @@ protected void preWrite(String instantTime,
WriteOperationType writeOperationTyp
@Override
protected void preCommit(String instantTime, HoodieCommitMetadata metadata) {
+ // Create a Hoodie table after startTxn which encapsulated the commits and
files visible.
+ // Important to create this after the lock to ensure latest commits show
up in the timeline without need for reload
+ HoodieTable<T, List<HoodieRecord<T>>, List<HoodieKey>, List<WriteStatus>>
table = createTable(config, hadoopConf);
+ TransactionUtils.resolveWriteConflictIfAny(table,
this.txnManager.getCurrentTransactionOwner(),
Review comment:
@leesf , IMO, the duplicate code is only duplicated with the
`SparkRDDWriteClient`, but the `HoodieJavaWriteClient` has no support to allow
parallel writing to tables using Locking service. Therefore in this pull
request, there is no need to move the duplicate code to super class. What's
more, @yihua could refactoring after this pull request, otherwise this has
conflict.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
