vinothchandar commented on a change in pull request #2374:
URL: https://github.com/apache/hudi/pull/2374#discussion_r588650200



##########
File path: 
hudi-client/hudi-client-common/src/main/java/org/apache/hudi/client/AbstractHoodieWriteClient.java
##########
@@ -359,10 +388,21 @@ public abstract O bulkInsertPreppedRecords(I 
preppedRecords, final String instan
    * Common method containing steps to be performed before write 
(upsert/insert/...
    * @param instantTime
    * @param writeOperationType
+   * @param metaClient
    */
-  protected void preWrite(String instantTime, WriteOperationType 
writeOperationType) {
+  protected void preWrite(String instantTime, WriteOperationType 
writeOperationType,
+      HoodieTableMetaClient metaClient) {
     setOperationType(writeOperationType);
-    syncTableMetadata();
+    
this.txnManager.setLastCompletedTransaction(metaClient.getActiveTimeline().getCommitsTimeline().filterCompletedInstants()
+        .lastInstant());
+    LOG.info("Last Instant Cached by writer with instant " + instantTime + " 
is " + this.txnManager.getLastCompletedTransactionOwner());
+    this.txnManager.setTransactionOwner(Option.of(new 
HoodieInstant(State.INFLIGHT, metaClient.getCommitActionType(), instantTime)));
+    this.txnManager.beginTransaction();
+    try {
+      syncTableMetadata();
+    } finally {
+      this.txnManager.endTransaction();

Review comment:
       I was wondering about the following scenario. if the cleanup fails, I 
guess it throws an error in both cases also? if so, this is okay. gtg




----------------------------------------------------------------
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.

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


Reply via email to