[Enhancement]: Speed up Search Iterator and Cleanup SDKs

[PwzXxm]

Is there an existing issue for this?

• I have searched the existing issues

What would you like to be added?

Speed up search iterators via using iterator directly rather than calling range search and maintain radius and range_filter on the SDK, which adds complexity to SDK implementations.

Why is this needed?

E2E Search Iterator becomes fundamental to workflows like client-side post filtering and others.

Anything else?

No response

[PwzXxm]

Stage 1 (planning to ship before 2.5.1)

1. Replace the underlying Range Search Calls with Knowhere iterator Next() calls. For each SDK Next() call, construct Knowhere iterators at segment level and calling Next() on them. NOT holding iterators among Next() calls at this stage.

• enhance: add search iterator v2 info milvus-proto#344
  • master enhance: speed up search iterator stage 1 #37947
  • 2.5 enhance: [2.5][cp] speed up search iterator stage 1 #38678
• SDKs (add another search_iterator_v2 API)
  • enhance: add search iterator v2 pymilvus#2395

Stage 2

1. Introduce SearchIteratorManager to hold iterators on each QN.
2. Limit the max number of iterator it can hold to limit the overall memory usage, modified LRU eviction.
3. Introduce iterator timeout settings and functionality.

Stage 3

1. Manually close search iterator, new gRPC call.
2. Calls sticky to delegators in multi-shard.

[PwzXxm]

/assign

[PwzXxm]

Stage 1 shipped with Milvus 2.5.2.
Stage 2 and 3 are on hold and pending further clarity.