micheal-o commented on code in PR #53386:
URL: https://github.com/apache/spark/pull/53386#discussion_r2632588138
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/StatePartitionWriter.scala:
##########
@@ -123,6 +159,12 @@ class StatePartitionAllColumnFamiliesWriter(
val valueRow = new
UnsafeRow(columnFamilyToValueSchemaLenMap(colFamilyName))
valueRow.pointTo(valueBytes, valueBytes.length)
- stateStore.put(keyRow, valueRow, colFamilyName)
+ if (columnFamilyToSchemaMap(colFamilyName).useMultipleValuesPerKey) {
+ // if a column family useMultipleValuesPerKey (e.g. ListType), we will
+ // write with 1 put followed by merge
+ stateStore.merge(keyRow, valueRow, colFamilyName)
Review Comment:
@zifeif2 merge alone won't work in all cases. The expectation is Put first
before merges
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
