>From Ali Alsuliman <[email protected]>:

Attention is currently required from: Peeyush Gupta, Vijay Sarathy, 
[email protected].
Ali Alsuliman has posted comments on this change. ( 
https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/19383 )

Change subject: [ASTERIXDB-3555][COMP] Use Join Samples to get Join Selectivity
......................................................................


Patch Set 6:

(2 comments)

File 
asterixdb/asterix-algebra/src/main/java/org/apache/asterix/optimizer/rules/cbo/Stats.java:

https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/19383/comment/67023c0e_f7ea6a81
PS5, Line 262: joinEnum.allJoinOps.get(0);
> Any join op node is fine. I  just need a join operator here. […]
Don't you need the "right" join operator? meaning if the join condition is 
coming from an INNER JOIN, shouldn't you make sure the join operator you are 
going to use is also an INNER JOIN? Getting the join operator at 0 may not 
necessarily be an INNER JOIN.
You can run this test and check:
    <test-case FilePath="tpch">
      <compilation-unit name="nest_aggregate">
        <output-dir compare="Text">nest_aggregate</output-dir>
      </compilation-unit>
    </test-case>


https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/19383/comment/b9b5dff5_831c18fe
PS5, Line 265: abjoin.getInputs().get(0)
> we dont need to switch back because we are making the change in a copy which 
> we do not need later on […]
It does not seem that this join operator is a copy. It's actually the same join 
operator coming from the plan. So, modifying it here is actually changing the 
plan. The join op inputs are switched back somehow late at the very end of the 
rule when the final tree is built. So, between this point here and the last 
point where the final tree is built, the join inputs are the sample indexes 
scan (not the data-scan operator). You want to check if having the sample 
indexes stick around as the join inputs during join enumeration is a problem or 
not.



--
To view, visit https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/19383
To unsubscribe, or for help writing mail filters, visit 
https://asterix-gerrit.ics.uci.edu/settings

Gerrit-Project: asterixdb
Gerrit-Branch: master
Gerrit-Change-Id: I863dc378d5ead9f8b48f09f7bc400f5bdaba4839
Gerrit-Change-Number: 19383
Gerrit-PatchSet: 6
Gerrit-Owner: [email protected]
Gerrit-Reviewer: Ali Alsuliman <[email protected]>
Gerrit-Reviewer: Anon. E. Moose #1000171
Gerrit-Reviewer: Jenkins <[email protected]>
Gerrit-Reviewer: Peeyush Gupta <[email protected]>
Gerrit-Reviewer: Vijay Sarathy <[email protected]>
Gerrit-Reviewer: [email protected]
Gerrit-Attention: Peeyush Gupta <[email protected]>
Gerrit-Attention: Vijay Sarathy <[email protected]>
Gerrit-Attention: [email protected]
Gerrit-Comment-Date: Fri, 07 Feb 2025 04:17:52 +0000
Gerrit-HasComments: Yes
Gerrit-Has-Labels: No
Comment-In-Reply-To: Ali Alsuliman <[email protected]>
Comment-In-Reply-To: [email protected]
Gerrit-MessageType: comment

Reply via email to