[enhancement](Nereids) fast compute hash code of deep expression tree to reduce conflict #38981
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
Contributor
Author
|
run buildall |
1 similar comment
Contributor
Author
|
run buildall |
924060929
changed the title
Contributor
Author
|
run buildall |
1 similar comment
Contributor
Author
|
run buildall |
TPC-H: Total hot run time: 41688 ms |
TPC-DS: Total hot run time: 168931 ms |
ClickBench: Total hot run time: 30.52 s |
Contributor
Author
|
run buildall |
924060929
force-pushed
the
fast_compute_hash_code
branch
from
896ebc7 to
b82026f
Compare
Contributor
Author
|
run buildall |
Contributor
Author
|
run buildall |
TPC-H: Total hot run time: 40712 ms |
TPC-H: Total hot run time: 40896 ms |
TPC-H: Total hot run time: 39884 ms |
TPC-H: Total hot run time: 39015 ms |
TPC-H: Total hot run time: 39338 ms |
TPC-H: Total hot run time: 39127 ms |
TPC-DS: Total hot run time: 203301 ms |
ClickBench: Total hot run time: 31.21 s |
morrySnow
approved these changes
Aug 8, 2024
github-actions
bot
added
the
approved
Contributor
|
PR approved by at least one committer and no changes requested. |
Contributor
|
PR approved by anyone and no changes requested. |
starocean999
approved these changes
Aug 8, 2024
924060929
added a commit
to 924060929/incubator-doris
that referenced
this pull request
Aug 9, 2024
… to reduce conflict (apache#38981) The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time. This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds (cherry picked from commit d5049ec)
924060929
added
dev/2.1.6-merged
dev/2.1.x
and removed
dev/2.0.x
dev/2.1.x
dev/2.1.6-merged
labels
924060929
added a commit
that referenced
this pull request
Aug 9, 2024
… to reduce conflict (#38981) (#39133) The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time. This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds
dataroaring
pushed a commit
that referenced
this pull request
Aug 11, 2024
… to reduce conflict (#38981) The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time. This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds
wyxxxcat
pushed a commit
to wyxxxcat/doris
that referenced
this pull request
Aug 14, 2024
… to reduce conflict (apache#38981) The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time. This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds
dataroaring
pushed a commit
that referenced
this pull request
Aug 16, 2024
… to reduce conflict (#38981) The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time. This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed changes
The Expression.hashCode default is getClass().hashCode(), just contains one level information, so the lots of expressions which is same type will return the same hash code and conflict, then compare deeply in the HashMap cause inefficient and hold table lock for long time.
This pr support fast compute hash code by the bottom literal and slot, reduce the compare expression time because of the conflict of hash code
In my test case, the sql planner time can reduce from 20 minutes(not finished) to 35 seconds