Change log

Features 🌰

• https://github.com/qdrant/qdrant/milestone/33?closed=1 - Inline Storage: Add option to inline vectors in HNSW graph for efficient IO usage (docs)
• https://github.com/qdrant/qdrant/milestone/37?closed=1 - Tenant promotion mechanism for tiered multitenancy: (docs)
• Add ReplicatePoints action to promote payload based tenant into dedicated shard key (docs)
• Add fallback shard key for intelligent routing to tenants that are or are not promoted to a dedicated shard (docs)
• On shard key creation, allow to specify initial state of new replicas
• https://github.com/qdrant/qdrant/pull/7414 - Add ACORN-1 search method, accurate search over many filtered points at the cost of performance (docs)
• https://github.com/qdrant/qdrant/pull/7408 - Add ASCII folding (normalization) to full text indices, fold diacritics into ASCII characters (docs)
• https://github.com/qdrant/qdrant/pull/7006 - Add conditional update functionality, only apply update on points matching filter (docs)
• https://github.com/qdrant/qdrant/pull/7100 - Add text_any full text filter to match any query term (docs)
• https://github.com/qdrant/qdrant/pull/7065 - Add option to customize RRF k parameter (docs)
• https://github.com/qdrant/qdrant/pull/7222 - In strict mode, specify maximum number of payload indices per collection
• https://github.com/qdrant/qdrant/pull/7123 - Add custom key-value metadata to collections (docs)
• https://github.com/qdrant/qdrant/pull/7291 - Add profiler to log slow point update and read requests

Improvements 🏎️

• https://github.com/qdrant/qdrant/pull/7385 - When loading Gridstore, populate tracker data into memory for faster first access
• https://github.com/qdrant/qdrant/pull/7407, https://github.com/qdrant/qdrant/pull/7405 - Spawn updates and flush workers task on updates runtime, use two system threads less for each local shard
• https://github.com/qdrant/qdrant/pull/7413 - Use system thread on demand in WAL, use one less system thread per local shard by default
• https://github.com/qdrant/qdrant/pull/7468 - Use atomic bit flags on HNSW construction to significantly increase indexing performance
• https://github.com/qdrant/qdrant/pull/7052, https://github.com/qdrant/qdrant/pull/7471 - Implement AVX512 SIMD optimizations for binary quantization on modern x86_64 CPUs
• https://github.com/qdrant/qdrant/pull/7433 - Switch new mutable payload indices and storage from RocksDB to Gridstore for better performance
• https://github.com/qdrant/qdrant/pull/7508 - Enable quantization in appendable segments by default, improving search performance
• https://github.com/qdrant/qdrant/pull/7347 - Change default score of query-less prefetch to 1.0 to ease score boosting
• https://github.com/qdrant/qdrant/pull/7369 - Don't explicitly disable strict mode by default
• https://github.com/qdrant/qdrant/pull/7345 - Simplify internal handling of copy-on-write segments, now write incoming updates to dedicated segments directly
• https://github.com/qdrant/qdrant/pull/7293, https://github.com/qdrant/qdrant/pull/7523 - Add warnings field to collection info, report misconfiguration
• https://github.com/qdrant/qdrant/pull/7319, https://github.com/qdrant/qdrant/pull/7401 - Report more helpful error messages on file IO errors
• https://github.com/qdrant/qdrant/pull/7377 - When using inference, propagate rate limit responses
• https://github.com/qdrant/qdrant/pull/7434 - Rate limit slow request warning to prevent spamming logs
• https://github.com/qdrant/qdrant/pull/7373 - Log shard transfers as a result of the consensus recovery procedure
• https://github.com/qdrant/qdrant/pull/7337 - Tone down optimizer logging to make it less verbose
• https://github.com/qdrant/qdrant/pull/7370 - Add TARGET_CPU and JEMALLOC_SYS_WITH_LG_PAGE build parameters to Docker image

Metrics 📈

• https://github.com/qdrant/qdrant/pull/7302, https://github.com/qdrant/qdrant/pull/7441 - In metrics, report point and vector counts per collection and vector name (collection_{points,vectors})
• https://github.com/qdrant/qdrant/pull/7307 - In metrics, report number of points skipped in indexed_only search (collection_indexed_only_excluded_points)
• https://github.com/qdrant/qdrant/pull/7301 - In metrics, report global effective minimum and maximum shard replication count (collection_active_replicas_{min,max})
• https://github.com/qdrant/qdrant/pull/7310, https://github.com/qdrant/qdrant/pull/7516 - In metrics, report total number of non-active replicas (collection_dead_replicas)
• https://github.com/qdrant/qdrant/pull/7316, https://github.com/qdrant/qdrant/pull/7480 - In metrics, expose collection_running_optimizations with number of optimizers running per collection
• https://github.com/qdrant/qdrant/pull/7497 - In metrics, report running, created and recovered number of snapshots (snapshot_{creation,recovery}_running, snapshot_created_total)
• https://github.com/qdrant/qdrant/pull/7484 - In metrics, report active thread count (process_threads)
• https://github.com/qdrant/qdrant/pull/7451 - In metrics, report number of open file descriptors and the limit (process_{open,max}_fds)
• https://github.com/qdrant/qdrant/pull/7487 - In metrics, report open mmaps and system limit (process_open_mmaps, system_max_mmaps)
• https://github.com/qdrant/qdrant/pull/7482 - In metrics, report total number of major and minor page faults (process_{minor,major}_page_faults_total)
• https://github.com/qdrant/qdrant/pull/7438 - In metrics, add configuration option to prefix all metrics with qdrant_ or something else

Bug fixes 🕵️

• https://github.com/qdrant/qdrant/pull/7527 - Fix logger API allowing arbitrary file writes, now this can only be configured through configuration
• https://github.com/qdrant/qdrant/pull/7530 - Abort pending search tasks when search is cancelled, fixing optimizer instability under huge load
• https://github.com/qdrant/qdrant/pull/7533 - Abort other blocking tasks such as retrieve and snapshot prematurely if the caller gets cancelled
• https://github.com/qdrant/qdrant/pull/7531 - Cancel ongoing searches more aggressively if the search is cancelled
• https://github.com/qdrant/qdrant/pull/7517 - Fix resharding down panic if no shard key is provided on collection with custom sharding
• https://github.com/qdrant/qdrant/pull/7469 - Fix panic on certain queries with unknown vector name
• https://github.com/qdrant/qdrant/pull/7375 - Forbid peer to join cluster with URI that is already used, which could break a cluster
• https://github.com/qdrant/qdrant/pull/7400 - Fix corrupt segments on load if segment was partially flushed, prevent payload index corruption
• https://github.com/qdrant/qdrant/pull/7342, https://github.com/qdrant/qdrant/pull/7404, https://github.com/qdrant/qdrant/pull/7427 - Force flush all segments when taking snapshot to prevent data corruption
• https://github.com/qdrant/qdrant/pull/7381 - Fix flush ordering on segments currently being snapshotted, fixing data consistency on crash
• https://github.com/qdrant/qdrant/pull/7388, https://github.com/qdrant/qdrant/pull/7416 - Fix flush ordering with concurrent flushes to ensure data consistency
• https://github.com/qdrant/qdrant/pull/7512, https://github.com/qdrant/qdrant/pull/7521 - Fix dummy shards not allowing snapshot recovery

Web UI 🫂

• https://github.com/qdrant/qdrant-web-ui/releases/tag/v0.2.0 - New web UI design to match Qdrant Cloud

Deprecations 🚧

• https://github.com/qdrant/qdrant/pull/7454 - Remove init_from collection API, deprecated since Qdrant 1.15
• https://github.com/qdrant/qdrant/pull/7449 - Remove lock API, deprecated since Qdrant 1.15
• https://github.com/qdrant/qdrant/pull/7047 - Remove old internal shard key format, deprecated and migrated away from in Qdrant 1.15
• https://github.com/qdrant/qdrant/pull/7450 - Remove payload filter from RBAC/JWT, deprecated since Qdrant 1.15, API keys using it are rejected
• https://github.com/qdrant/qdrant/pull/7183 - Deprecate old variant of vector output