Skip to content

Add on_disk_rescore to checkvec tool #137593

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.

Sign up for GitHub

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
elasticsearchmachine merged 1 commit into from
Nov 4, 2025
Merged

Add on_disk_rescore to checkvec tool #137593

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 12 additions & 2 deletions qa/vector/src/main/java/org/elasticsearch/test/knn/CmdLineArgs.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,8 @@ record CmdLineArgs(
KnnIndexTester.MergePolicyType mergePolicy,
double writerBufferSizeInMb,
int writerMaxBufferedDocs,
int forceMergeMaxNumSegments
int forceMergeMaxNumSegments,
boolean onDiskRescore
) implements ToXContentObject {

static final ParseField DOC_VECTORS_FIELD = new ParseField("doc_vectors");
Expand Down Expand Up @@ -90,6 +91,7 @@ record CmdLineArgs(
static final ParseField MERGE_POLICY_FIELD = new ParseField("merge_policy");
static final ParseField WRITER_BUFFER_MB_FIELD = new ParseField("writer_buffer_mb");
static final ParseField WRITER_BUFFER_DOCS_FIELD = new ParseField("writer_buffer_docs");
static final ParseField ON_DISK_RESCORE_FIELD = new ParseField("on_disk_rescore");

/** By default, in ES the default writer buffer size is 10% of the heap space
* (see {@code IndexingMemoryController.INDEX_BUFFER_SIZE_SETTING}).
Expand Down Expand Up @@ -135,6 +137,7 @@ static CmdLineArgs fromXContent(XContentParser parser) throws IOException {
PARSER.declareDouble(Builder::setWriterBufferMb, WRITER_BUFFER_MB_FIELD);
PARSER.declareInt(Builder::setWriterMaxBufferedDocs, WRITER_BUFFER_DOCS_FIELD);
PARSER.declareInt(Builder::setForceMergeMaxNumSegments, FORCE_MERGE_MAX_NUM_SEGMENTS_FIELD);
PARSER.declareBoolean(Builder::setOnDiskRescore, ON_DISK_RESCORE_FIELD);
}

@Override
Expand Down Expand Up @@ -209,6 +212,7 @@ static class Builder {
private long seed = 1751900822751L;
private KnnIndexTester.MergePolicyType mergePolicy = null;
private double writerBufferSizeInMb = DEFAULT_WRITER_BUFFER_MB;
private boolean onDiskRescore = false;

/**
* Elasticsearch does not set this explicitly, and in Lucene this setting is
Expand Down Expand Up @@ -360,6 +364,11 @@ public Builder setForceMergeMaxNumSegments(int forceMergeMaxNumSegments) {
return this;
}

public Builder setOnDiskRescore(boolean onDiskRescore) {
this.onDiskRescore = onDiskRescore;
return this;
}

public CmdLineArgs build() {
if (docVectors == null) {
throw new IllegalArgumentException("Document vectors path must be provided");
Expand Down Expand Up @@ -397,7 +406,8 @@ public CmdLineArgs build() {
mergePolicy,
writerBufferSizeInMb,
writerMaxBufferedDocs,
forceMergeMaxNumSegments
forceMergeMaxNumSegments,
onDiskRescore
);
}
}
Expand Down
3 changes: 2 additions & 1 deletion qa/vector/src/main/java/org/elasticsearch/test/knn/KnnIndexTester.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -130,7 +130,8 @@ static Codec createCodec(CmdLineArgs args) {
format = new ESNextDiskBBQVectorsFormat(
encoding,
args.ivfClusterSize(),
ES920DiskBBQVectorsFormat.DEFAULT_CENTROIDS_PER_PARENT_CLUSTER
ES920DiskBBQVectorsFormat.DEFAULT_CENTROIDS_PER_PARENT_CLUSTER,
args.onDiskRescore()
);
} else if (args.indexType() == IndexType.GPU_HNSW) {
if (quantizeBits == 32) {
Expand Down