Skip to content

[8.0] [MOD-9557] Fix incorrect vector blob size calculation #672

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 11, 2025

Conversation

meiravgri
Copy link
Collaborator

backport #665 to 8.0
conflict resoultions were related to svs files.

* add     size_t bg_vector_indexing_count;
    size_t main_vector_indexing_count;

* fix datasize in executeInsertJob

* disable neighbours print

* replace manual blob size calcaulations

format

comment out new tests

* INT8Test::PopulateRandomVector
increases seed every call so we will get different vectors.
this required a change in metrics_test because now the score is not exactly 0.

VecSimIndexAbstract CastToBruteForce()-> rename to GetFlatBufferIndex

add BruteForceIndex CastToBruteForce()

add getFlatBufferIndexAsBruteForce to tiered index to get BruteForceIndex

functions added to hnsw and BF:
Formalize getDataByLabel definition: populate the intput empty vector with the vectors data,not including any additional meta data that might present in the vector storage like the norm.
add getStoredVectorDataByLabel returns a vector of char vectors. each vector contains the vector as it is stored in the vectors container, including the meta data if exists.

* build  test with VERBOSE=1

* verbose cov

* verbose cov fix

assert in FP32SpacesOptimizationTest::FP32InnerProductTest

* fix

* add force_copy to maybeCopyToAlignedMem and preprocessQuery

use preprocessQuery with force_copy in batch iterator instead on preprocessQueryInPlace

* also batch in svs

* revert redundancies

* tiered batch iterator doesn't copy the blob

* add
                // TODO: handle original_blob_size != processed_bytes_count

* revert pull request changes

* move getDataByLabel and getStoredVectorDataByLabel to VecSimIndexAbstract

* add test_index_test_utils
currently testing get_stored_vector_data_single_test

* getFlatBufferIndex returns BruteForceIndex

* little test name fix

* small fix

* fix range

(cherry picked from commit 98cf0a1)
@meiravgri meiravgri requested review from alonre24 and GuyAv46 and removed request for alonre24 May 11, 2025 08:55
@meiravgri meiravgri enabled auto-merge May 11, 2025 08:56
Copy link

codecov bot commented May 11, 2025

Codecov Report

Attention: Patch coverage is 95.71429% with 3 lines in your changes missing coverage. Please review.

Project coverage is 96.71%. Comparing base (173df56) to head (c10882a).
Report is 1 commits behind head on 8.0.

Files with missing lines Patch % Lines
src/VecSim/algorithms/hnsw/hnsw_multi.h 85.71% 2 Missing ⚠️
.../VecSim/algorithms/brute_force/brute_force_multi.h 90.00% 1 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##              8.0     #672      +/-   ##
==========================================
- Coverage   96.92%   96.71%   -0.21%     
==========================================
  Files         107      107              
  Lines        5729     5761      +32     
==========================================
+ Hits         5553     5572      +19     
- Misses        176      189      +13     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

@meiravgri meiravgri added this pull request to the merge queue May 11, 2025
Merged via the queue into 8.0 with commit c9259dc May 11, 2025
16 of 17 checks passed
@meiravgri meiravgri deleted the backport-665-to-8.0 branch May 11, 2025 11:25
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants