maedhroz commented on a change in pull request #1083:
URL: https://github.com/apache/cassandra/pull/1083#discussion_r661005855
##########
File path: src/java/org/apache/cassandra/db/SinglePartitionReadCommand.java
##########
@@ -1033,7 +1033,15 @@ private ClusteringIndexNamesFilter
reduceFilter(ClusteringIndexNamesFilter filte
SearchIterator<Clustering, Row> searchIter =
result.searchIterator(columnFilter(), false);
- PartitionColumns columns = columnFilter().fetchedColumns();
+ // According to the CQL semantics a row exists if at least one of its
columns is not null (including the primary key columns).
+ // Having the queried columns not null is unfortunately not enough to
prove that a row exists has some column deletion
+ // for the queried columns can exist on another node.
+ // For CQL tables it is enough to have the primary key liveness and
the queried columns as the primary key liveness prove that
+ // the row exists even if all the other columns are deleted.
+ // COMPACT tables do not have primary key liveness by consequence we
are forced to get all the fetched columns to ensure that
Review comment:
```suggestion
// COMPACT tables do not have primary key liveness, and by
consequence we are forced to get all the fetched columns to ensure that
```
--
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]
