veghlaci05 commented on code in PR #3746:
URL: https://github.com/apache/hive/pull/3746#discussion_r1021470141
##########
ql/src/java/org/apache/hadoop/hive/ql/txn/compactor/CompactionQueryBuilder.java:
##########
@@ -287,16 +302,27 @@ private void buildAddClauseForAlter(StringBuilder query) {
private void buildSelectClauseForInsert(StringBuilder query) {
// Need list of columns for major crud, mmmajor partitioned, mmminor
List<FieldSchema> cols;
- if (major && crud || major && insertOnly && sourcePartition != null ||
minor && insertOnly) {
+ if (rebalance || major && crud || major && insertOnly && sourcePartition
!= null || minor && insertOnly) {
if (sourceTab == null) {
return; // avoid NPEs, don't throw an exception but skip this part of
the query
}
cols = sourceTab.getSd().getCols();
} else {
cols = null;
}
-
- if (crud) {
+ if (rebalance) {
+ query.append("0, t2.writeId, t2.rowId % ");
+ query.append(numberOfBuckets);
Review Comment:
Yes, in order to equally distribute the records among the buckets, the
numberOfBuckets % row_number is used.
##########
ql/src/java/org/apache/hadoop/hive/ql/txn/compactor/CompactionQueryBuilder.java:
##########
@@ -287,16 +302,27 @@ private void buildAddClauseForAlter(StringBuilder query) {
private void buildSelectClauseForInsert(StringBuilder query) {
// Need list of columns for major crud, mmmajor partitioned, mmminor
List<FieldSchema> cols;
- if (major && crud || major && insertOnly && sourcePartition != null ||
minor && insertOnly) {
+ if (rebalance || major && crud || major && insertOnly && sourcePartition
!= null || minor && insertOnly) {
if (sourceTab == null) {
return; // avoid NPEs, don't throw an exception but skip this part of
the query
}
cols = sourceTab.getSd().getCols();
} else {
cols = null;
}
-
- if (crud) {
+ if (rebalance) {
+ query.append("0, t2.writeId, t2.rowId % ");
Review Comment:
The first value id the 'operation'. In other compactions it is set by the
validate_acid_sort_order UDF, but in the rebalancing case that function would
be a more difficult to use.
--
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]
