[GitHub] [cassandra] adelapena commented on pull request #1739: CASSANDRA-8877 trunk: Ability to read the TTL and WRITE TIME of an element in a collection

[GitBox <gi...@apache.org>]

adelapena commented on PR #1739:
URL: https://github.com/apache/cassandra/pull/1739#issuecomment-1198066349

   > How is the forward compatibility? For example, the cluster is in the mixed mode and the old node (as the coordinator) receives the query with WRITETIME and it further queries replicas running both old and new versions.
   
   In that case, if the column is not a collection the old coordinator would run the query as usual. Nothing has changed for regular columns so the new nodes would return the same. If the column is a collection the old coordinator would reject the query before that wasn't supported until now.
   
   As for queries running on the new node, they would work even if there are not upgraded nodes, because everything happens on the coordinator's CQL layer. 
   
   The problematic case would be aggregation queries such as `MAX(WRITETIME)` or `MAX(TTL)` running on a new node while there are not-upgraded nodes. Those queries won't work properly because the selector is serialized and sent to the not-upgraded nodes. So I think we should add a note advising against using these functions on not-frozen collections/UDTs when upgrading to 4.2, similar to the one that we added for `MAXWRITETIME`. 


-- 
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: pr-unsubscribe@cassandra.apache.org

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


---------------------------------------------------------------------
To unsubscribe, e-mail: pr-unsubscribe@cassandra.apache.org
For additional commands, e-mail: pr-help@cassandra.apache.org