LakshSingla commented on code in PR #16800:
URL: https://github.com/apache/druid/pull/16800#discussion_r1755197654
##########
server/src/main/java/org/apache/druid/server/ClientQuerySegmentWalker.java:
##########
@@ -840,6 +846,7 @@ private static <T, QueryType extends Query<T>>
Optional<DataSource> materializeR
+ "from the query context and/or the server
config."
);
} else {
+ resultSequence.set(results);
return Optional.empty();
Review Comment:
> your answer suggests that the situation have changed
Yes, the fallback cases are mostly known and preventable.
> couldn't it be changed to fallback early and don't even try?
I think it would require a change in the tool chests since there isn't a
method currently to get the row signature while specifying the finalization.
Well, essentially `resultArraySignature` would deem the aggregator's signature
as finalization = UNKNOWN even though we want the finalization = YES.
##########
server/src/main/java/org/apache/druid/server/ClientQuerySegmentWalker.java:
##########
@@ -840,6 +846,7 @@ private static <T, QueryType extends Query<T>>
Optional<DataSource> materializeR
+ "from the query context and/or the server
config."
);
} else {
+ resultSequence.set(results);
return Optional.empty();
Review Comment:
> your answer suggests that the situation have changed
Yes, now the fallback cases are mostly known and preventable.
> couldn't it be changed to fallback early and don't even try?
I think it would require a change in the tool chests since there isn't a
method currently to get the row signature while specifying the finalization.
Well, essentially `resultArraySignature` would deem the aggregator's signature
as finalization = UNKNOWN even though we want the finalization = YES.
--
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]
