saurabhd336 commented on code in PR #10172:
URL: https://github.com/apache/pinot/pull/10172#discussion_r1086375466
##########
pinot-core/src/main/java/org/apache/pinot/core/query/executor/ServerQueryExecutorV1Impl.java:
##########
@@ -475,6 +476,10 @@ private static void getSegmentExplainPlanRowData(Operator
node, ExplainPlanRows
}
}
+ if (node instanceof AcquireReleaseColumnsSegmentOperator) {
+ ((AcquireReleaseColumnsSegmentOperator) node).materializeChildOperator();
Review Comment:
Are you saying this because the run() method for some PlanNode can read
actual buffer to generate the operator?
While that is true (case in point:
https://github.com/apache/pinot/blob/master/pinot-core/src/main/java/org/apache/pinot/core/plan/FilterPlanNode.java#L291),
prefetch / acquire will only make whole thing faster by fetching those buffers
everything in advance right? Even if we refer to the buffers without prefetch /
acquire, it'll still work (albeit slow)? I was able to get correct plan outputs
with both the SegmentLocalFSDirectory (where prefetch / acquire are NO-OP), and
some other implementations of the SegmentDirectory where these functions are
implemented.
I guess the only time this will fail is if some SegmentDirectory
implementation necessarily needs prefetch / acquire before being able to access
the buffers at all?
--
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]
