SearchIteratorAsync()
Creates a server-side iterator that pages over search results in batches.
public SearchIterator SearchIteratorAsync(SearchIteratorReq request)
Request Syntax
var request = new SearchIteratorReq
{
CollectionName = "my_collection",
VectorFieldName = "embedding",
Vectors = [new[] { 0.1f, 0.2f, 0.3f, 0.4f }],
MetricType = MetricType.L2,
Limit = 100,
BatchSize = 20
};
SearchIterator iterator = client.SearchIteratorAsync(request);
REQUEST PARAMETERS:
CollectionName(string) -The name of the collection to search in.
DatabaseName(string?) -An optional database name to operate on. When empty, the client’s currently selected database is used.
VectorFieldName(string) -The name of the vector field to search in.
Vectors(IReadOnlyList> ) -The query vectors to search for (dense float vectors). Only a single query vector is supported.
SparseVectors(IReadOnlyList>? ) -The sparse query vectors to search for. When set,
Vectorsmust be empty. See SparseVector.Float16Vectors(IReadOnlyList>? ) -The float16 query vectors to search for. When set,
Vectorsmust be empty.BFloat16Vectors(IReadOnlyList>? ) -The bfloat16 query vectors to search for. When set,
Vectorsmust be empty.BinaryVectors(IReadOnlyList>? ) -The binary query vectors to search for (raw bytes). When set,
Vectorsmust be empty.Int8Vectors(IReadOnlyList>? ) -The int8 query vectors to search for. When set,
Vectorsmust be empty.Texts(IReadOnlyList? ) -The text query strings to search for (embedded-text / BM25 targets). When set,
Vectorsmust be empty.EmbeddingLists(IReadOnlyList? ) -The embedding lists to search for, used for Array-of-Struct vector sub-fields such as
"clips[vector]"with a MAX_SIM metric. The search iterator accepts a single query, so provide exactly one embedding list. See EmbeddingList.MetricType(MetricType) -The metric type used to measure the distance between vectors. See MetricType.
Limit(int) -The maximum total number of rows to return.
0orint.MaxValuemeans no limit.Parameters(SearchParameters?) -The optional search parameters. See SearchParameters.
BatchSize(int) -The number of rows to fetch per batch. Defaults to 1000.
ExternalFilterFunc(Func>? ) -An optional client-side per-page filter that receives each page as a SingleResult and returns the zero-based row indices to keep, in the desired order. Filtered rows are accumulated and returned in batches of
BatchSize, with a final partial batch when no more results remain.
RETURN TYPE:
SearchIterator
RETURNS:
A SearchIterator that pages over search results as an async iterator of batches. It uses the search_iter_v2 server protocol and requires a Milvus server of version 2.5.2 or later. Consume with await foreach, optionally with .WithCancellation(token).
The SearchIterator exposes the following members:
METHODS:
GetAsyncEnumerator(CancellationToken cancellationToken = default)-Returns an async enumerator that yields each page of search results as a SingleResult holding that page’s top-K hits with their scores, primary keys and output fields.
EXCEPTIONS:
-
Thrown when the RPC fails or the server returns an error. Inspect the exception message for details.
Example
using Milvus.Client.V2;
using Milvus.Client.V2.Requests.Dql;
using Milvus.Client.V2.Types;
using var client = new MilvusClientV2(new ConnectConfig
{
Uri = "localhost:19530",
Username = "root",
Password = "Milvus"
});
await client.ConnectAsync();
await foreach (SingleResult batch in client.SearchIteratorAsync(new SearchIteratorReq
{
CollectionName = "my_collection",
VectorFieldName = "embedding",
Vectors = [new[] { 0.1f, 0.2f, 0.3f, 0.4f }],
MetricType = MetricType.L2,
Limit = 100,
BatchSize = 20
}))
{
// batch holds the page's top-K scores, primary keys and output fields.
}