impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Thomas Tauber-Marshall (Code Review)" <>
Subject [Impala-ASF-CR] IMPALA-4252: Min-max runtime filters for Kudu
Date Thu, 26 Oct 2017 20:52:06 GMT
Thomas Tauber-Marshall has posted comments on this change. (

Change subject: IMPALA-4252: Min-max runtime filters for Kudu

Patch Set 8:


> I noticed that - BloomFilterBytes is always 0 in the query
 > profiles.
 > Can you please veirfy?

Yes. This line is always printed to the profile, even if there are no bloom filters being
built. We don't expose the mem used by min-max filters in the profile as its negligible.
Commit Message:
PS7, Line 26: 
> Might be worth mentioning why the runtime filters were renumbered in all th
File be/src/exec/filter-context.h:
PS7, Line 106:   Status CloneFrom(const FilterContext& from, ObjectPool* pool, RuntimeState*
> Long lines
File be/src/util/min-max-filter.h:
PS7, Line 140:   virtual void* GetMax() override { return &max_; }
> Nit: not sure why this is using underscores while AlwaysTrue() is using cam
PS7, Line 176:   StringValue min_;
> Can you briefly mention what it means when these are empty - that the value
File be/src/util/
PS6, Line 291:     out->__isset.min = true;
> Oh ok, thanks for clarifying. It looked superficially like a last line copy
File be/src/util/
PS7, Line 137:     DCHECK(thrift.__isset.max);
> Do we have end-to-end tests for these code paths? I think we could generall
I've now added e2e and unit tests for all of the truncation scenarios.
PS7, Line 143:     CopyToBuffer(&max_buffer_, &max_, max_.len);
> This has a fairly large number of edge cases - it would be good to unit tes
PS7, Line 143: uff
> static_cast instead of int()?
PS7, Line 150: 
> I would have expected that we would modify the trailing values that overflo
PS7, Line 207:     out->min.__set_string_val(in.min.string_val);
> It seems tricky to test these various error paths in end-to-end tests. Coul
As you say, its difficult to test the oom scenario in an e2e test, since the max amount of
mem being used here is so small, but its covered with a unit test now.

To view, visit
To unsubscribe, visit

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I02bad890f5b5f78388a3041bf38f89369b5e2f1c
Gerrit-Change-Number: 7793
Gerrit-PatchSet: 8
Gerrit-Owner: Thomas Tauber-Marshall <>
Gerrit-Reviewer: Anonymous Coward #345
Gerrit-Reviewer: Lars Volker <>
Gerrit-Reviewer: Matthew Jacobs <>
Gerrit-Reviewer: Michael Ho <>
Gerrit-Reviewer: Mostafa Mokhtar <>
Gerrit-Reviewer: Thomas Tauber-Marshall <>
Gerrit-Reviewer: Tim Armstrong <>
Gerrit-Reviewer: Todd Lipcon <>
Gerrit-Comment-Date: Thu, 26 Oct 2017 20:52:06 +0000
Gerrit-HasComments: Yes

  • Unnamed multipart/alternative (inline, 8-Bit, 0 bytes)
View raw message