* weighted rrf implementation
* test
* fmt
* fix edge
* validate number of sources and number of weights
* do not partial match
* upd schema
* review fixes
* update formula
* remove calcualtions from tests
* update comment, because AI have OCD
* fmt
Shutdown `shard` explicitly at the end of benchmark to avoid error
```
A Tokio 1.x context was found, but it is being shutdown.
```
Co-authored-by: Ivan Boldyrev <ivan.boldyrev@qdrant.com>
* allow custom K parameter for RRF
* generate grpc docs and openapi
* use tagged type approach for parametrized fusions
* use params approach in grpc
* simplify api structure
* upd schema
* nits
* rest: parameterized rrf as query variant
* consistency in doc comments
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* Retain more closed segments
* Add back rocksdb
* Update WalOptions/WalConfig across the code
* Use NonZeroUsize
* Expose via APIs
* Update gRPC docs
* Fix stoarge compat test
* recompile openapi.json with rocksdb
* default wal retain closed fn
* update openapi.json
* Use qdrant/wal latest commit and remove from config.yaml
* Deprecate init_from
* Mark memmap_threshold as deprecated
* Mark locks API as deprecated
* Mark RBAC collection access payload filter as deprecated
* Allow deprecations in tests and benchmarks
* Bump Rust edition to 2024
* gen is a reserved keyword now
* Remove ref mut on references
* Mark extern C as unsafe
* Wrap unsafe function bodies in unsafe block
* Geo hash implements Copy, don't reference but pass by value instead
* Replace secluded self import with parent
* Update execute_cluster_read_operation with new match semantics
* Fix lifetime issue
* Replace map_or with is_none_or
* set_var is unsafe now
* Reformat
* rename cpu_budget -> resource_budget
* clippy
* add io budget to resources
* fmt
* move budget structures into a separate file
* add extend permit function
* dont extend existing permit
* switch from IO to CPU permit
* do not release resource before aquiring an extension
* fmt
* Review remarks
* Improve resource permit number assertion
* Make resource permit replace_with only acquire extra needed permits
* Remove obsolete drop implementation
* allocate IO budget same as CPU
* review fixes
---------
Co-authored-by: timvisee <tim@visee.me>
* bump and migrate to rand 0.9.0
also bump rand_distr to 0.5.0 to match it
* Migrate AVX2 and SSE implementations
* Remove unused thread_rng placeholders
* More random migrations
* Migrate GPU tests
* bump seed
---------
Co-authored-by: timvisee <tim@visee.me>
Co-authored-by: Arnaud Gourlay <arnaud.gourlay@gmail.com>
* Add UUID to collection...
...and recreate collection, when applying Raft snapshot, if UUID of collection is different
* fixup! Add UUID to collection...
Remove UUID field from gRPC and exclude it from OpenAPI spec 🤡
* fixup! fixup! Add UUID to collection...
Always generate collection UUID 🤦♀️
* Raft snapshot recreate collection no expose UUID (#5452)
* separate colleciton config structure from API
* fmt
* Update lib/collection/src/operations/types.rs
Co-authored-by: Tim Visée <tim+github@visee.me>
---------
Co-authored-by: Tim Visée <tim+github@visee.me>
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
Co-authored-by: Tim Visée <tim+github@visee.me>
* Revert "Experiment: in stream records, set cutoff to latest clocks receiver is guaranteed to have (#5375)"
This reverts commit e843647c95.
* Revert "Set replica state function does not need to be async anymore (#5379)"
This reverts commit fdf08e0a59.
* Revert "Experiment: disable clocks in initializing state, propagate ignore flag (#5372)"
This reverts commit 0d98f6fe1a.
* Revert "Experiment: also don't write clock tags to WAL in partial state (#5352)"
This reverts commit f69ec61848.
* Revert "Experiment: ignore clock tags when replica is in partial state (#5349)"
This reverts commit 1b8a38f70b.
* Keep change to not send updates to replicas in recovery state
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* Propagate flag for ignoring clocks on local shard from replica set
* Also ignore local clocks in initializing state, is similar to partial
* Remove previous logic for disabling clocks
* We can make static replica set state functions inlined const
* Fix typo
* include document & image objects into grpc API
* introduce image and object to rest api
* minor refactoring
* rename Vector -> VectorInternal
* decompose vector data structures
* add schema
* fmt
* grpc docs
* fix conversion
* fix clippy
* fix another conversion
* rename VectorInput -> VectorInputInternal
* replace grpc TryFrom with async functions
* fmt
* replace rest TryFrom with async functions
* add image and object into query rest
* separate inference related conversions
* move json-related conversions into a separate file
* move vector-related transformations into a separate file
* move more vector related-conversions into dedicated module
* Propagate payload index schema down to shard replica set + update handler
* Configure payload indices when creating new appendable segment
* When loading segments, make sure applied payload indices match config
* Add test to assert creating new segments with payload index
* Fix unit test because the collection payload schema wasn't updated
* Add test for updating payload index configuration on segment load
* Update test documentation
* Also create payload indices in temporary snapshot segment
* do not delete extra payload index from segments
* do not delete extra payload index from segments
* fix test
---------
Co-authored-by: generall <andrey@vasnetsov.com>
* minor conversion improvement
* use NamedVectors in update_vectors
* remove merge from VectorStruct
* rename Multi -> Named in vector struct
* add multi-dense vectors option into VectorStruct
* generate openapi
* rename VectorStruct -> VectorStructInternal
* add conversion for anonymous multivec in grpc
* renames for BatchVectorStruct
* implement multi-dense for batch
* allow multi-dense in batch upserts
* test and fixes
* universal-query: add with_lookup to query API definition
* fix consistency doc check
* lookup_from and not lookup_with
* clean import
* improve doc for lookup location wrt using
* Suggestions for descriptions
---------
Co-authored-by: Luis Cossío <luis.cossio@outlook.com>
* Move CPU count function to common, fix wrong CPU count in visited list
* Change default number of rayon threads to 8
* Use CPU budget and CPU permits for optimizer tasks to limit utilization
* Respect configured thread limits, use new sane defaults in config
* Fix spelling issues
* Fix test compilation error
* Improve breaking if there is no CPU budget
* Block optimizations until CPU budget, fix potentially getting stuck
Our optimization worker now blocks until CPU budget is available to
perform the task.
Fix potential issue where optimization worker could get stuck. This
would happen if no optimization task is started because there's no
available CPU budget. This ensures the worker is woken up again to
retry.
* Utilize n-1 CPUs with optimization tasks
* Better handle situations where CPU budget is drained
* Dynamically scale rayon CPU count based on CPU size
* Fix incorrect default for max_indexing_threads conversion
* Respect max_indexing_threads for collection
* Make max_indexing_threads optional, use none to set no limit
* Update property documentation and comments
* Property max_optimization_threads is per shard, not per collection
* If we reached shard optimization limit, skip further checks
* Add remaining TODOs
* Fix spelling mistake
* Align gRPC comment blocks
* Fix compilation errors since last rebase
* Make tests aware of CPU budget
* Use new CPU budget calculation function everywhere
* Make CPU budget configurable in settings, move static budget to common
* Do not use static CPU budget, instance it and pass it through
* Update CPU budget description
* Move heuristic into defaults
* Fix spelling issues
* Move cpu_budget property to a better place
* Move some things around
* Minor review improvements
* Use range match statement for CPU count heuristics
* Systems with 1 or 2 CPUs do not keep cores unallocated by default
* Fix compilation errors since last rebase
* Update lib/segment/src/types.rs
Co-authored-by: Luis Cossío <luis.cossio@qdrant.com>
* Update lib/storage/src/content_manager/toc/transfer.rs
Co-authored-by: Luis Cossío <luis.cossio@qdrant.com>
* Rename cpu_budget to optimizer_cpu_budget
* Update OpenAPI specification
* Require at least half of the desired CPUs for optimizers
This prevents running optimizations with just one CPU, which could be
very slow.
* Don't use wildcard in CPU heuristic match statements
* Rename cpu_budget setting to optimizer_cpu_budget
* Update CPU budget comments
* Spell acquire correctly
* Change if-else into match
Co-authored-by: Luis Cossío <luis.cossio@qdrant.com>
* Rename max_rayon_threads to num_rayon_threads, add explanation
* Explain limit in update handler
* Remove numbers for automatic selection of indexing threads
* Inline max_workers variable
* Remove CPU budget from ShardTransferConsensus trait, it is in collection
* small allow(dead_code) => cfg(test)
* Remove now obsolete lazy_static
* Fix incorrect CPU calculation in CPU saturation test
* Make waiting for CPU budget async, don't block current thread
* Prevent deadlock on optimizer signal channel
Do not block the optimization worker task anymore to wait for CPU budget
to be available. That prevents our optimizer signal channel from being
drained, blocking incoming updates because the cannot send another
optimizer signal. Now, prevent blocking this task all together and
retrigger the optimizers separately when CPU budget is available again.
* Fix incorrect CPU calculation in optimization cancel test
* Rename CPU budget wait function to notify
* Detach API changes from CPU saturation internals
This allows us to merge into a patch version of Qdrant. We can
reintroduce the API changes in the upcoming minor release to make all of
it fully functional.
---------
Co-authored-by: Luis Cossío <luis.cossio@qdrant.com>
Co-authored-by: Luis Cossío <luis.cossio@outlook.com>
* remove duplicated search methods, introduced for compatibility in last version
* Use `with_capacity` rather than a manual reserve
* explicit Arc clones
* get rid of batching by runs of same strategy
* avoid refactor in group by
* more explicit arc clones, remove one .expect()
* little extra refactor on recommendations.rs
* refactor grouping_test.rs too
---------
Co-authored-by: timvisee <tim@visee.me>
Co-authored-by: Luis Cossío <luis.cossio@outlook.com>
* add timeout query param for search requests
* enable timeout for recommend requests
* Add query timeout for group by requests
* update openapi models
* Don't decrease timeout after recommend preprocessing
* Add openapi test
* code review
* add timeout to individual group by requests, non-decreasing
* handle timeout for discover
* Update timeout field tag in SearchBatchPoints
message
This is particularly useful for microbenchmarks since
1) this cost shouldn't affect benchmark results
2) memory allocations are the largest source of measurement noise
* Clone current segment config to deprecated type
* Remove segment level quantization config from segment config
* Also deprecate current VectorDataConfig
* Update old segment migration to work with new refactoring
* Move index into vector data config
* Move vector data config migration logic into segment level
* Remove hnsw_config from vector data config
* Rename collection params to vector data conversions function
* Move storage type into vector data config
* Set appendable flag correctly
* Clean up and reformat
* Make segment on disk flag not optional
* Add appendable flag to segment config to replace storage type
* Remove storage type from segment config
* Deprecate storage type enum
* Use consistent variable naming
* Cleanup
* Add segment config migration for v0.5.0 to current
* Bump segment to 0.6.0
* Remove serde defaults for new storage and vector data config types
These default value configurations are not needed anymore, because these
structs are not used to deserialize old data. All current fields should
always be available in these structs. When new fields are added in new
functions, the serde default annotation must be set again.
* Cleanup
* Update OpenAPI specification
This updates the returned data structure on telemetry endpoints, as a
result of segment configuration refactoring.
* Fix quantization configuration not falling back to collection config
* Fix compiler warning when building in release mode
* Move deprecated type structs into compat module
* Update allow deprecated attributes
* Assign quantization config only in segment optimizer
* Remove unsued parameter
* Add vector storage type enum to vector data config
* Remove appendable and on_disk flags from segment and vector config
* Update OpenAPI specification
* add tests
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
* wip: chunked mmap
* Fix typo
* insert and get methods
* dynamic bitvec
* clippy
* wip: vector storage
* wip: fmt
* wip: mmap chunks
* wip: mmap problems
* Share transmuted mutable reference over mmap
* option to enable appendable mmap vectors
* fmt
* rename storage status file
* update tests
* fix get deleted value range
* add recovery to vector storage tests
* add flush to tests
* fix transmute from immutable to mutable
* make transmuted pointer private
* remove unused unsafe functions
* force WAL flush if wait=true
* move wal flush into updater thread
* remove flush from update api
* Minimize pub visibility for specialized/dangerous functions
* Allocate vector with predefined capacity
* Inline format parameters
* Assert we have multiple chunks while testing, test is useless otherwise
* Remove unnecessary scope
* Remove unnecessary dereference
* Random bool has 0.5 as standard distribution, use iter::repeat_with
* Replace RemovableMmap::new with Default derive
* Rename len to num_flags
* Use Option replace as it is convention alongside take
* Add FileId enum to replace error prone manual ID rotating
* Use debug_assert_eq where applicable
* Refactor drop and set to replace
* Change default chunk size for chunked mmap vectors to 32MB
This change is made as per GitHub review, because allocating a few
storages with 128MB would take a significant amount of time and storage.
See: https://github.com/qdrant/qdrant/pull/1838#discussion_r1187215475
* Replace for-loops with iterators
* Draft: add typed mmap to improve code safety (#1860)
* Add typed mmap
* Replace some crude mmap usages with typed mmap
* Use typed mmap for deleted flags
* Simplify dynamic mmap flags a lot with new typed mmap, remove flags option
* Reformat
* Remove old mmap functions that are now unused
* Reimplement mmap locking for mmap_vectors
* Add MmapBitSlice tests
* Replace MmapChunk with new typed mmap
* Update docs
* Clean-up
* Disable alignment assertions on Windows for now
* Rename mmap lock to mlock to prevent confusion with lockable types
* one more small test
* Some review fixes
* Add aliasing note
* Add basic error handling in typed mmap constructors
* Use typed mmap error handling throughout project
* Move mmap type module to common
* Fix transmute functions being unsound
See https://github.com/qdrant/qdrant/pull/1860#discussion_r1188593854
---------
Co-authored-by: Andrey Vasnetsov <andrey@vasnetsov.com>
---------
Co-authored-by: timvisee <tim@visee.me>
Co-authored-by: Tim Visée <tim+github@visee.me>
