* Don't insert deferred points into sparse index
# Conflicts:
# lib/segment/src/segment/tests.rs
# lib/segment/src/segment_constructor/segment_constructor_base.rs
# Conflicts:
# lib/segment/src/segment_constructor/segment_constructor_base.rs
* Clippy
* Assert consistency of deferred_internal_id var
* Return before SparseVector conversion in case of deferred point
* Use debug_assert instead
* Refactor: `SearchSegmentEntry` for search ops
* Move `has_deferrend_points_method`
to `SearchSegmentEntry`
* Reorg imports
* Renamings per review
* fixup! Renamings per review
* `StorageSegmentEntry` trait
It contains methods dealing with storage and syncronization
* Move index methods from `ReadSegmentEntry`
to `SegmentEntry`
* Add `_concurrent` suffix to some methods
* move field index methods into NonAppendableSegmentEntry
---------
Co-authored-by: generall <andrey@vasnetsov.com>
* UIO traits: require Sized
I don't think we ever going to add unsized implementation. So, lets,
require it on the whole trait rather than on individual methods.
* Move `AccessPattern` to the `common` package
Was: segment::vector_storage::vector_storage_base::AccessPattern
Now: common::generic_consts::AccessPattern
* UIO traits: replace `bool` with `AccessPattern`
Instead of `IdTracker` to have `iter_*` methods, it now has a `point_mappings()` method, returning a guard to `PointMapping` value. It prepares for moving the point_mappings under a lock to allow parallel searches and deletions.
Also, an incorrect unsafe in `Segment::iter_points` is replaced by a safe version with `self_cell`.
* feat(universal_io): add storage-agnostic BitSliceStorage with read/write support`
Add BitSliceStorage<S> generic over
UniversalRead<u64>/UniversalWrite<u64>,
providing bit-level read and write operations over u64-element storage.
Read operations (S: UniversalRead<u64>):
- read_all: returns Cow<BitSlice> (zero-copy for mmap, owned for others)
- get_bit: single bit read via u64 element fetch
- read_bit_range: arbitrary bit range read
Write operations (S: UniversalWrite<u64>):
- set_bit / replace_bit: single bit write (skips write if unchanged)
- write_bit_range: arbitrary bit range write from BitSlice source
- fill_bit_range: fill range with a value
- set_bits_batch: batch individual bit updates coalesced by element
- flusher: flush underlying storage
* refactor: replace MmapBitSlice with BitSliceStorage in MmapBitSliceBufferedUpdateWrapper
Migrate all deleted-flag bitslice storage from the legacy MmapBitSlice
(Deref-based mmap wrapper) to BitSliceStorage<MmapUniversal<u64>>
(storage-agnostic universal IO backend).
Updated consumers:
- MmapMapIndex
- MmapNumericIndex
- MmapGeoMapIndex
- MmapInvertedIndex (fulltext)
- ImmutableIdTracker
- Benchmark
Additionally:
- Add BitSliceStorage::create(path, num_bits) to encapsulate
file creation + sizing + open (eliminates duplicated size math
across 5 call sites)
- Add MmapBitSliceStorage type alias for
BitSliceStorage<MmapUniversal<u64>>
- Add count_ones() convenience method
- Use set_bits_batch() in wrapper flusher and all build paths
instead of per-bit set_bit() loops (coalesces u64 read-modify-writes)
- Fix silent error swallowing in wrapper get(): log + debug_assert
on I/O errors instead of .ok().flatten()
- Remove dead bitmap_mmap_size function from immutable_id_tracker
* use bitvec's approach for offset
* less api
* misc improvements
* refactor write method
* move to segment crate
* handle creation of file outside of StoredBitSlice logic
* fix codespell
* document bitwise calculations
* coalesce more updates into a single write, batch all writes
* use native `extend_from_bitslice` instead of iterating.
* clarity refactor
* clippyyy
* use `u64` as BitStore everywhere
* assume iterator is sorted
* only use chunk_by for generating runs
* fix update wrapper flusher
* review fixes
* larger set bits batch test
* remove reintroduced file
* oops, fix new test
---------
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: generall <andrey@vasnetsov.com>
* Add deferred point count to UpdateQueueInfo
* Hide if prevent_unoptimized is false
* Clippy
* Openapi
* Iterate over appendable segments only
* Only calculate deferred point count if prevent_unoptimized is true
* Rebase fixes
* Codespell
* Don't account for deferred points in some places
# Conflicts:
# lib/collection/src/shards/local_shard/scroll.rs
* Add in QueryContext
* Cover more places
* Coderabbit review remarks
* Properly count amount of deleted deferred points (#8386)
* Properly count amount of deleted deferred points
* Prevent double-counting of the same point
* Remove hints to estimations
* Properly handle counts in ProxySegment
* Add tests and fix deleted point count issue
* Adjusts tests + fix issues
* Separte fields for deferred points in telemetry (SegmentInfo)
* Remove deferred_points_count()
* openapi
* Fix test by manually calculating visible points
* Adjust ProxySegment test to revertion of SegmentInfo
* Throw error if collection was not found in telemetry (e2e Test)
* Update lib/segment/src/segment/segment_ops.rs
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* Make new fields in API optional
* Don't take range if no deferred point exist
* Review remarks
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* Log individual payload index load time per field
Add per-field timing to `load_all_fields()` in `StructPayloadIndex`,
using the existing `LOAD_TIMING_LOG_TARGET` infrastructure. Each
payload field index that takes >= 5ms to load now gets its own log
line, making it easy to identify slow-loading fields.
Made-with: Cursor
* Move `log_load_timing` to common module and use everywhere
Move the `log_load_timing` helper from `segment_constructor_base` into
`common::defaults` so it can be reused across crates. Convert all raw
`log::debug!(target: LOAD_TIMING_LOG_TARGET, ...)` call sites to use
the shared function, giving consistent formatting and min-duration
suppression everywhere.
Made-with: Cursor
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
Reduce test parameters across 5 areas to cut test runtime without
sacrificing coverage:
1. HNSW PQ tests: lower vector dimensionality from 131 to 64 for product
quantization variants, cutting PQ codebook training time (~31s -> ~8s)
2. HNSW index build: use 2 threads instead of 1 for index construction;
the tests only check accuracy above a threshold so determinism is not
required
3. Search attempts: reduce from 10 to 5 query vectors per test
4. Rescoring: skip the expensive re-upsert-all-as-zeros rescoring check
for PQ tests (already covered by scalar quantization variants)
5. Near-miss speedups:
- WAL: reduce QuickCheck iterations from 100 to 50
- Gridstore: halve operation counts, drop 64-byte block size case,
reduce proptest cases for gap search
- Continuous snapshot: reduce timeout from 20s to 10s
Made-with: Cursor
Co-authored-by: Cursor Agent <agent@cursor.com>
* Replace `has_deferred_points` with `deferred_points_count` in SegmentEntry
Change the `SegmentEntry` trait method from returning a boolean to
returning the actual count of deferred points (`usize`). This gives
callers more information without needing a separate method, while
existing boolean checks simply compare against zero.
Made-with: Cursor
* Fix clippy collapsible_if warning in deferred_points_count
Made-with: Cursor
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
Add a `log_load_timing` helper that checks elapsed time against a 5ms
threshold before logging. Sub-component loads faster than this round to
"0.00s" in the {:.2} format and are pure noise. The per-segment "total
loaded" line remains unconditional.
Made-with: Cursor
Co-authored-by: Cursor Agent <agent@cursor.com>
Add `deferred_internal_id` field to `SegmentInfo` so it is exposed
through segment telemetry. The field is optional and skipped from
serialization when `None`. It is excluded from anonymization via
`#[anonymize(false)]`.
Made-with: Cursor
Co-authored-by: Cursor Agent <agent@cursor.com>
* Improvements for `IoUringFile`
- handle read and write requests more explicitly
- assert against partial reads/writes
* Implement `read_multi`/`write_multi`
* Implement `populate` and `clear_ram_cache`
* Rename internal structures as `IoUringSomething`
* [ai] Make `UniversalRead`/`UniversalWrite` impl generic over `T`
* Handle `io_uring` initialization error
* Check that `io_uring` is initialized and supported, when opening `IoUringFile`
* fixup! Check that `io_uring` is initialized and supported, when opening `IoUringFile`
Fix typo
* test for reading u64 from file with uring
* fmt
* clippy
* fix(io_uring): allocate Vec<MaybeUninit<T>> for reads to fix alignment for T (#8353)
* fix(io_uring): allocate Vec<MaybeUninit<T>> for reads to fix alignment for T
Refactor IoUringState::read to take generic T and item_offset/items_length
instead of byte_offset/byte_length. Allocate Vec<MaybeUninit<T>> so the
kernel writes into correctly aligned memory, then convert to Vec<T> in
finalize. Fixes bytemuck::cast_vec alignment panic for types like u64.
Made-with: Cursor
* refactor(io_uring): only make read/write methods generic, not state or runtime
- IoUringState and IoUringRuntime no longer generic over T
- read<T>() allocates Vec<MaybeUninit<T>>, transmutes to Vec<MaybeUninit<u8>> for storage
- finalize returns ReadBuffer; callers use .into_vec::<T>() to get Vec<T>
- Write path unchanged (callers pass bytes via bytemuck::cast_slice)
Made-with: Cursor
* Cleanup
* review n1
* review n2
* fmt
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
Co-authored-by: Roman Titov <ffuugoo@users.noreply.github.com>
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* use AHash instead of just hash
* add ahash to deps
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
Co-authored-by: qdrant-cloud-bot <111755117+qdrant-cloud-bot@users.noreply.github.com>
Co-authored-by: Cursor Agent <agent@cursor.com>
* Add debug-level load timing logs for storage components
Add log::debug! calls with target "qdrant::load_timing" that report
how long each storage component takes to load during startup:
- Total shard load time (including WAL replay)
- Total segment load time + load_state
- Per-segment: payload_storage, id_tracker, payload_index
- Per-vector: vector_storage (dense/sparse), quantized_vectors,
vector_index (dense/sparse)
These logs are at DEBUG level so they are silent by default. Enable
selectively via log_level config or env, for example:
log_level: "INFO,qdrant::load_timing=debug"
No config fields, no global flags, no function signature changes.
Made-with: Cursor
* Use seconds with 10ms resolution for load timing logs, extract log target constant
Switch from `{:.2}ms` with `as_secs_f64() * 1000.0` to `{:.2}s` with
`as_secs_f64()` for cleaner output. Move the "qdrant::load_timing"
log target string to a LOAD_TIMING_LOG_TARGET constant in each file.
Made-with: Cursor
* Move LOAD_TIMING_LOG_TARGET to common::defaults
Single definition shared by segment and collection crates, avoiding
duplication of the log target string.
Made-with: Cursor
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
* chore: gitignore .idea and .vscode globally
Reason: `lib/edge/publish` is a separate Cargo workspace and I'd like to
have a separate rust-analyzier configuration for it, which is stored in
`.vscode/settings.json`. Not sure about `.idea`, but I guess the same
logic applies.
* chore: Use Iterator::is_sorted (stabilized in Rust 1.82)
* chore: remove unrelated files
These were added in #5501 (42fd2e27), perhaps by accident?
* Add ignore_deferred flag for internal scroll API
* Ignore deferred points in ProxyShard::update
* Use enum instead of bool as type in function parameters
* Use consistent parameter name
* fmt
* Apply suggestions from code review
Co-authored-by: Tim Visée <tim+github@visee.me>
* More explicit naming
* Add DeferredBehavior to `count()` and disable filtering in stream_records
* Don't filter `retrieve()` everywhere
* Use consistent filter behavior in read operations
* Add TODO for ignored parameter in remote_shard
* Rebase fixes
---------
Co-authored-by: timvisee <tim@visee.me>
Co-authored-by: Tim Visée <tim+github@visee.me>
* Change backpressure policy for deferred points
* back pressure waits for deferred points
* Remove prevent_unoptimized_threshold_kb
* tests
* are you happy fmt
* better optimization wait
* fix after rebase
* send signal if optimizer is over the budget
* Make wait_for_deferred_points_ready cancel safe
* fix test
---------
Co-authored-by: Ivan Pleshkov <pleshkov.ivan@gmail.com>
Co-authored-by: timvisee <tim@visee.me>
* Implement filtering: Search
* Add tests for search and read_filtered
* Implement filtering: read_ordered_filtered + Tests
* Codespell + Clippy
* Reduce code duplication in tests
* Use helper methods of Id tracker
* Implement filtering: Random Scroll
* fixes after rebase
* [Unittest] Only set deferred ID if deferred points exist
* [Unit test] Reduce combinations in segment creation test
---------
Co-authored-by: Ivan Pleshkov <pleshkov.ivan@gmail.com>
* [manunal] points deduplication after optimization
* take a list of deferred point ids
* test
* review remarks
* review remarks
---------
Co-authored-by: Ivan Pleshkov <pleshkov.ivan@gmail.com>
* Use array_windows when statically sized windows were used
* Bump MSRV to 1.94
* fix in gpu code
---------
Co-authored-by: Luis Cossío <luis.cossio@outlook.com>
* Adjust points selection for deferred points update
* adjust proxy segment implementation
* simplify
* use simpler proxy impl
* stick to Entry API
* renaming to stay closer to the original
* two passes and simpler impl.
* fmt
* fmt
* use predefined deferred internal id
* calculate deferred point id
* move deferred check to the entry
* fix after rebase
* fmt
* fix tests
* review remarks
* fix tests
* codespell fix
* are you happy clippy
---------
Co-authored-by: Arnaud Gourlay <arnaud.gourlay@gmail.com>
* Introduce EdgeShardConfig for edge shard
- Add EdgeShardConfig and EdgeOptimizersConfig in lib/edge/src/config.rs
- Segment config (vector_data, sparse_vector_data, payload_storage_type)
- Global hnsw_config and per-vector HNSW in segment config
- Optimizer params: deleted_threshold, vacuum_min_vector_number,
default_segment_number, max_segment_size, indexing_threshold,
prevent_unoptimized (excludes memmap_threshold, flush_interval_sec,
max_optimization_threads)
- Persist/load as edge_config.json in shard path
- EdgeShard uses RwLock<EdgeShardConfig>; load() accepts Option<EdgeShardConfig>,
falls back to file or infer from segments; compatibility checked on load
- load_with_segment_config() for backward compatibility (SegmentConfig -> EdgeShardConfig)
- optimize() uses EdgeShardConfig for hnsw and optimizer thresholds
- Public methods: set_hnsw_config(), set_vector_hnsw_config(), set_optimizers_config()
(update and persist)
- Python and examples use load_with_segment_config with existing config API
Made-with: Cursor
* Refactor EdgeShardConfig: user-facing params only, config module
- Replace SegmentConfig inside EdgeShardConfig with user-facing fields:
- on_disk_payload (bool) instead of payload_storage_type
- vectors: HashMap<VectorNameBuf, EdgeVectorParams> with on_disk per vector,
no per-vector quantization; global quantization_config only
- sparse_vectors: HashMap<VectorNameBuf, EdgeSparseVectorParams> with on_disk
- EdgeVectorParams / EdgeSparseVectorParams use on_disk (bool) instead of
storage_type; conversion to VectorDataConfig/SparseVectorDataConfig in
to_segment_config()
- Add config module: mod.rs, optimizers.rs, vectors.rs, shard.rs
- from_segment_config(&SegmentConfig) fills all inferrable params
- to_segment_config() builds SegmentConfig for segments and optimize()
- load_with_segment_config takes Option<SegmentConfig>, uses from_segment_config
Made-with: Cursor
* Move optimizer threshold helpers to shard crate
- Add get_number_segments, get_indexing_threshold_kb, get_max_segment_size_kb,
get_deferred_points_threshold_bytes in shard::optimizers::config
- Collection OptimizersConfig and edge EdgeOptimizersConfig delegate to these
- Single place for threshold logic; collection and edge use shard helpers
Made-with: Cursor
* Use destructuring in config conversions to avoid missing new fields
- EdgeVectorParams: destructure VectorDataConfig in from_*, destructure self in to_vector_data_config
- EdgeSparseVectorParams: destructure SparseVectorDataConfig and SparseIndexConfig in from_*, destructure self in to_sparse_vector_data_config
- EdgeShardConfig: destructure SegmentConfig in from_segment_config, destructure self in to_segment_config
Adding new fields to source structs will now cause compile errors until conversions are updated.
Made-with: Cursor
* refactor: centralize on_disk_payload→payload_storage_type, on_disk→storage_type, and appendable quantization logic
- PayloadStorageType::from_on_disk_payload(bool) in segment (Mmap/InRamMmap)
- VectorStorageType::from_on_disk(bool) in segment (ChunkedMmap/InRamChunkedMmap)
- QuantizationConfig::for_appendable_segment(Option<&Self>) in segment (feature flag + supports_appendable)
- collection: use from_on_disk_payload in non-rocksdb branch
- edge shard/vectors: use new helpers; remove duplicated conditionals
- shard optimizers: use from_on_disk and for_appendable_segment
Made-with: Cursor
* refactor(edge): use EdgeShardConfig directly, drop segment_config
- Add plain_segment_config() for create_appendable_segment (no HNSW)
- Add segment_optimizer_config() built from EdgeShardConfig for blocking optimizers
- Add vector_data_config(name) for query/MMR
- build_blocking_optimizers: use segment_optimizer_config() instead of SegmentConfig
- create_appendable_segment: use plain_segment_config()
- search/query: use config().vectors and vector_data_config() instead of segment_config()
- Remove segment_config() from EdgeShardConfig and EdgeShard
- Add to_plain_vector_data_config on EdgeVectorParams
Made-with: Cursor
* [manual] review changes
* refactor(edge-py): wrap EdgeShardConfig, add EdgeVectorParams/EdgeSparseVectorParams
- PyEdgeConfig now wraps EdgeShardConfig (vectors, sparse_vectors, on_disk_payload, etc.)
- PyEdgeVectorParams / PyEdgeSparseVectorParams wrap edge config types
- PyEdgeOptimizersConfig for optional optimizer settings
- EdgeShard.load() uses EdgeShardConfig; edge::config made pub for Python crate
- cargo fmt + clippy (remove map_identity)
Made-with: Cursor
* refactor(edge-py): simplify config API, remove unused Py* types, add EdgeConfig
- Remove unused PyPayloadStorageType, PyVectorDataConfig, PyVectorStorageType,
PySparseVectorDataConfig, PySparseVectorStorageType from Python bindings
- Move PyEdgeOptimizersConfig to lib/edge/python/src/config/optimizers.rs
- Update qdrant_edge.pyi: EdgeConfig with vectors/sparse_vectors,
EdgeVectorParams, EdgeSparseVectorParams, EdgeOptimizersConfig
- Update examples (common.py, repr.py) to use new config API
- Run cargo fmt
Made-with: Cursor
* [manual] review changes
* [manual] review changes
* [manual] fix test
* Address CodeRabbit review comments for PR 8322 (#8324)
* Address CodeRabbit review comments for PR 8322
- Python examples: explicit imports (repr.py, common.py) and new EdgeConfig API
- HnswIndexConfig: add max_indexing_threads param and property in .pyi and Rust bindings
- EdgeConfig: make vectors optional for sparse-only configs; validate at least one of vectors/sparse_vectors
- EdgeShardConfig::load: use try_exists(), propagate I/O errors
- from_segment_config: infer hnsw_config from per-vector HNSW when all agree
- EdgeShard setters: atomic clone-mutate-save-then-replace; persist config save errors
- Segment compat: prefix vector name in error messages; resolve None datatype to Float32
- max_indexing_threads: preserve 0 (auto) sentinel in trait default; remove per-optimizer overrides
- SegmentOptimizerConfig:🆕 build plain and optimizer maps in single pass
- config_mismatch_optimizer tests: use VectorNameBuf::from() instead of .into()
- vectors.rs: doc updates for per-vector quantization
Made-with: Cursor
* Address @generall review: SaveOnDisk for config, resolve num_rayon_threads in optimizer
- Use SaveOnDisk<EdgeShardConfig> for EdgeShard config (generall: 'We have SaveOnDisk struct for this')
- Create via SaveOnDisk::new() after resolving config; setters use .write() for atomic persist
- set_vector_hnsw_config: clone then mutate then write (fallible setter)
- max_indexing_threads: resolve 0 (auto) via num_rayon_threads inside impl (generall: 'proper solution would be to resolve num_rayon_threads inside the optimizer impl')
- max_indexing_threads_sentinel_aware() now returns Some(num_rayon_threads(raw)) so callers get actual thread count
Made-with: Cursor
* [manual] reorganize num_rayon_threads -> get_num_indexing_threads to better account per-vector configuration
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
Co-authored-by: generall <andrey@vasnetsov.com>
* update docstring and pyi
* fmt
* fmt
* clipy
---------
Co-authored-by: Cursor Agent <agent@cursor.com>
Co-authored-by: generall <andrey@vasnetsov.com>
* Adjust points selection for deferred points update
* adjust proxy segment implementation
* clean
* simplify
* use simpler proxy impl
* cleanup
* stick to Entry API
* renaming to stay closer to the original
* renaming to stay closer to the original
* nits
* two passes and simpler impl.
* fmt
* fmt
---------
Co-authored-by: Ivan Pleshkov <pleshkov.ivan@gmail.com>
