Redis Open Source 8.12 release notes

Redis Open Source 8.12 release notes.

This section is deprecated, and its content will be moved or removed in a future documentation update.

Redis Open Source 8.12-RC1 (October 2026)

This is the first Release Candidate of Redis 8.12 in Redis Open Source.

Release Candidates are feature-complete pre-releases. Pre-releases are not suitable for production use.

Headlines:

Redis 8.12 introduces new features and performance improvements.

Operating systems we test Redis 8.12 on

Ubuntu 22.04 (Jammy Jellyfish), 24.04 (Noble Numbat), 26.04 (Resolute Raccoon)
Rocky Linux 8.10, 9.7, 10.1
AlmaLinux 8.10, 9.7, 10.1
Debian 12.13 (Bookworm), Debian 13.4 (Trixie)
Alpine 3.23
macOS 14.8.4 (Sonoma), 15.7.4 (Sequoia), 26.3 (Tahoe) - for both Intel and ARM

Security fixes (compared to 8.10.0)

#Q10073 Indexed value hashing is collision-resistant, preventing index degradation from hash-based denial-of-service (HashDoS) attacks. (MOD-14459)
#Q11202 unicode_tolower no longer reads out of bounds on a dangling multi-byte lead byte in indexed TEXT fields encoded with the 8-bit Unicode Transformation Format (UTF-8). (MOD-18088)

New Features (compared to 8.10)

#15331 Compressed bitmaps - redis-roaring integration reduces memory usage and speeds up bitmap operations. Thanks to Antonio, its creator, for contributing this capability.
#15986 New command: BITRANGEINTARRAY - returns set bit offsets within a range, equivalent to redis-roaring's R.RANGEINTARRAY.
#15938 Set expiration on individual set members:
- New commands: SEXPIRE, SPEXPIRE, SEXPIREAT, SPEXPIREAT - set expiration.
- New commands: STTL, SPTTL, SEXPIRETIME, SPEXPIRETIME - retrieve remaining time-to-live (TTL) values or expiration timestamps.
#15828 Bless keys: avoid eviction of specified keys; new command: BLESS.
#15829 CLUSTER MIGRATION STATUS reports progress for both the snapshot and incremental migration phases.
#15556 Configure Transport Layer Security (TLS) named groups with tls-groups.
#P1063 Count-min sketches support optional CELL_SIZE in CMS.INITBYDIM and CMS.INITBYPROB for lower memory usage. CMS.INCRBY accepts negative increments to decrement counts. CMS.INFO reports cell sizes.
#J1640 Automatically create deep paths in JavaScript Object Notation (JSON).
#Q11330 8-bit scalar quantization (SQ8) compression is available for hierarchical navigable small world (HNSW) vector fields. (MOD-14958)
#Q10063 COLLECT LIMIT honors the configurable search-max-aggregate-results.

Bug fixes (compared to 8.10.0)

#15569 Changes made by XDELEX DELREF for already-deleted stream entries could fail to propagate to replicas or the append-only file (AOF).
#15508 Operations using random radix-tree traversal could behave incorrectly when starting from certain leaf nodes.
#15604 Memory usage could be calculated incorrectly for hashes using shared templates.
#15489 Redis database (RDB) files could fail to load after XSETID reduced a stream's entries_added counter.
#15626 SET could calculate an incorrect expiration time when given an excessively large relative TTL.
#15628 Active defragmentation could incorrectly handle the shared-template registry used by compact hashes.
#15530 After failover, a replica could remain connected to another replica instead of the new primary.
#15733 Invalid cluster slot ranges could be accepted after being merged into a valid range.
#15756 RESTORE could calculate an incorrect expiration time when given an excessively large relative TTL.
#15699 Memory growth could be accounted for incorrectly when SETBIT or BITFIELD expanded a string.
#15815 XTRIM could leave a stream in a state that prevented an RDB file containing it from loading.
#15795 Clients could fail to receive a protocol error when a malformed command arrived across multiple network reads.
#15875 ARLASTITEMS could omit elements appearing after a gap in an array.
#15916 Data could be lost when a legacy AOF file was upgraded before loading completed.
#15917 Hiredis clients on Linux could mishandle errors when establishing Transmission Control Protocol (TCP) connections.
#15869 Redis could crash when a module thread paused input/output (I/O) threads.
#15468 Publish/subscribe (Pub/Sub) subscriptions could remain active after the subscribing user's channel permissions were revoked or the user was deleted, if the client had re-authenticated as another user.
#15623 Module actions scheduled by hash-field expiration notifications could be delayed after fields expired in the background.
#15680 Data could be lost after a replica migrated to a different shard.
#15587 LMOVEM could emit keyspace notifications inconsistent with those produced by LMOVE.
#15695 Command side effects could be propagated to replicas or the AOF despite settings that should have excluded them.
#Q10872 The shard no longer crashes with a segmentation fault (SIGSEGV) when RESTORE of a legacy ft_index0 payload is issued outside an RDB load. (MOD-15685)
#Q10728 Arming the garbage collection (GC) timer off the main thread on ARM no longer causes random shard crashes after extended uptime. (MOD-17309)
#Q10767 The server no longer crashes on writes when an index FILTER references no schema fields. (MOD-15592)
#Q11303, #Q11304, #Q11308, #Q11309, #Q11310, #Q11311 FT.ALTER SCHEMA ADD running concurrently with queries under WORKERS>0 no longer crashes query execution across VECTOR, TAG, NUMERIC, GEO, ISMISSING, and GEOSHAPE nodes. (MOD-18356, MOD-18361, MOD-18366, MOD-18367, MOD-18368)
#Q11474 FT.AGGREGATE with the numeric SORTBY optimizer no longer crashes or returns wrong results. (MOD-6967)
#Q10209 The coordinator no longer crashes with Redis serialization protocol version 2 (RESP2) when FT.PROFILE HYBRID hits a return-strict timeout. (MOD-16396)
#Q10145 TEXT PHONETIC matches now behave correctly when non-TEXT fields precede TEXT fields in the schema. (MOD-16304)
#Q11262 An FT.AGGREGATE cursor with FILTER reports the correct result count. (MOD-18172)
#Q11282 FT.EXPLAIN now renders a multi-word tag value as a single tag rather than an INTERSECT. (MOD-8666)
#Q10768 NOT queries no longer leak excluded documents or drop results after a concurrent index change. (MOD-17394)
#Q10637 Numeric index cardinality and bounds are now correct under float compression. (MOD-16877)
#Q10704 Empty wildcard patterns no longer cause out-of-bounds writes or spurious empty-value matches. (MOD-17451)
#Q10752 Wildcard ? on WITHSUFFIXTRIE TEXT fields now matches per codepoint rather than per byte. (MOD-17476)
#Q10112 Suffix-trie wildcard queries on TEXT fields are now case-insensitive. (MOD-16259)
#Q11450 Wildcard_MatchRune no longer hangs on a literal * in the indexed string.
#Q11267 Duplicate vector results are no longer returned during tiered indexing. (MOD-17795)
#Q10731 FT.SEARCH on a cluster returns an error for invalid LIMIT arguments, matching standalone behavior. (MOD-12465)
#Q10251 FT.SEARCH under WORKERS>0 no longer returns a nil field-array for documents re-indexed mid-query. (MOD-16507)
#Q10921 Sort keys with embedded NUL bytes are no longer truncated in replies.
#Q10085 The RETURN_STRICT safe-loader no longer deadlocks the Redis global lock (GIL) on client timeout. (MOD-16267)
#Q11449 An index reference leak after queued search cancellation no longer occurs. (MOD-18682)
#Q11369 Memory leaks after coordinator SEARCH cancellation no longer occur. (MOD-18548)
#Q10881 SST replication no longer hangs and the numeric snapshot is no longer corrupted when disk GC is mid-cycle. (MOD-17197)
#Q10110 FT.HYBRID cursor setup now has a bounded channel timeout. (MOD-16145)
#Q10211, #Q10745 RETURN_STRICT cursors are depleted on timeout; strict-timed-out hybrid cursor cycles drop their cursors at cycle end. (MOD-16091, MOD-16991)
#Q10947 Coordinator cursor memory is now capped via INDEX_CURSOR_LIMIT. (MOD-17479)
#Q10766, #Q10763 Cursor reads no longer time out against an earlier read's deadline and correctly report a timeout that expires while iterators are revalidated. (MOD-17489, MOD-17488)
#Q10993 RDB and replication payloads no longer leak a stray heap byte. (MOD-17800)
#Q10074 FT.HYBRID returns a stable reciprocal rank fusion (RRF) result order across cluster shards. (MOD-12438)
#Q11027 Cluster TLS connections now use Redis node metadata, aligning with topology changes. (MOD-15869)
#Q10380, #Q10430 Cluster topology is now refreshed on topology-change events rather than by polling. (MOD-15868, MOD-15870)
#Q10835 Coordinator fanout now preserves binary argument lengths.
#Q11095 Replica stalls during FT.DROPINDEX DD are reduced.
#Q10083 Internal query arguments missing during a rolling upgrade are tolerated. (MOD-16047)
#Q10816 Forced GC now waits for a concurrent child process instead of skipping its cycle.
#Q10958 The asynchronous disk read pool no longer spins when saturated; it blocks instead.
#Q10646 The VectorSimilarity dependency addresses unaligned metadata access on ARM/Scalable Vector Extension (SVE) paths. (MOD-15303)
#Q10352 FT.PROFILE ON_TIMEOUT FAIL now reports a pre-execution timeout as a warning rather than an error. (MOD-16660)
#Q10870 Crash reports now include in-flight coordinator aggregate, hybrid, and cursor-read queries. (MOD-16992)
#Q11089 Debug aggregate request synchronization no longer crashes. (MOD-17616)
#J1636 Fixed array-index normalization overflow for extreme negative indexes, including i64::MIN, affecting commands such as JSON.ARRTRIM.
#J1648 Corrected JSON.MGET key-spec metadata to identify all key arguments and exclude the trailing JSONPath argument.
#J1670 Fixed downstream compilation when using redis_json_module_create! with the as-library feature and the new RDB depth configuration.
#P1035 TDIGEST.QUANTILE, TDIGEST.CDF, and TDIGEST.TRIMMED_MEAN could return a not-a-number (NaN) value for a sketch with exactly one observation.
#P1053 RDB loading and RESTORE inconsistently rejected invalid numeric values, allocation sizes, serialized counts, or inconsistent metadata in Bloom filters, Count-Min Sketches, Top-K sketches, and t-digests. Bloom filter growth lacked overflow checks.
#P1067 Bloom filters with excessive hash-function counts could be accepted during loading and restoration, including through BF.LOADCHUNK.
#P1053 Top-K sketches containing nonempty items with zero counts or empty-string items with nonzero counts could be restored incorrectly.
#P1062 CMS.MERGE key metadata did not correctly identify the destination and all source keys, affecting cluster slot validation.
#P1071 CF.COMPACT command metadata was missing its argument count and key specification.

Performance and resource utilization improvements (compared to 8.10.0)

#15618 Batch updates to hashes using shared templates run more efficiently.
#15018 Supported x86 systems use the processor-based monotonic clock by default, with runtime calibration to reduce timing overhead.
#15782 RDB and AOF loading provides sequential-read hints to the operating system to improve loading efficiency.
#Q10169 Searches on indexes using field-level expiration (HEXPIRE) are faster. (MOD-16350)
#Q10981, #Q11065 Coordinator field lookup and row-reply serialization overhead is reduced. (MOD-17924)
#Q11055, #Q11056, #Q11057 Cluster reply conversion, result destruction, and GROUPBY cleanup have reduced overhead. (MOD-17908, MOD-17909, MOD-17910)
#Q11507, #Q11506 Query replies now format integral numeric values without snprintf and declare reply collection lengths up front. (MOD-18800, MOD-18799)
#Q11509 Hash field loading for RETURN and LOAD no longer requires per-field allocations. (MOD-18811)
#Q11278 FT.ALTER avoids unnecessary document reindexing during schema additions. (MOD-18171)
#Q11384 Score/payload-only hash updates avoid reindexing. (MOD-18182)
#Q11230 Hash documents are no longer re-indexed when no indexed field is touched. (MOD-17688)
#Q10515 Coordinator requests are now allocated on the main thread, removing CoordRequestCtx. (MOD-16990)
#Q11018 Query usage is now separated from administrative index operations. (MOD-17536)
#J1645 Memory usage reporting divides a shared string's allocation, including header and padding, proportionally among its references. Previously, each reference was charged the full allocation.
#J1660 JSON parsing uses compact object allocation to reduce memory usage.
#J1661 JSONPath parsing and evaluation reduce backtracking, temporary allocations, and repeated traversal. Optimizations cache root-path operands across filter candidates, stop existence checks at the first match, and reduce regex matching and string comparison overhead.
#J1631 Improved typed-array element access through direct indexing, avoiding repeated iteration and quadratic work during full indexed scans.
#P1049 Improved t-digest compression performance for inputs containing many duplicate values. Sorting is bounded to O(n log n), avoiding quadratic behavior on pathological inputs.

Configuration parameters

#15331 redis-roaring: bitmap-default-roaring - enables the new roaring bitmaps type; disabled by default.
#J1657 json-legacy-rdb-depth-limit enforces a maximum nesting depth of 128 when loading JSON from RDB or through RESTORE. It is disabled by default.
#J1640 json-auto-create-deep-paths is disabled by default and configurable only at load time. Use the same value across shard nodes and AOF replays to prevent data divergence.

Metrics

#15374 INFO keysizes histograms include stream key sizes.
#15682 INFO STREAMS provides per-database histograms of consumer groups per stream, pending entries per group, and consumers per group. Enable with stream-stats yes (disabled by default).
#Q10097 FT.INFO now exposes the default EF_RUNTIME. (MOD-16147)
#Q10107 Disk GC bytes_collected and cycle statistics are reported in FT.INFO. (MOD-14621)
#Q10464 FT.INFO on the coordinator aggregates per-field disk metrics via the reducer. (MOD-14859)
#Q10204 Redis INFO now breaks down query timeout metrics by pipeline stage. (MOD-13192)
#Q11521 Index metrics are no longer collected for unrelated INFO sections. (MOD-18033)
#Q10536 FT.PROFILE loader-field timing resolution is improved. (MOD-17132)

Behavior changes

#Q10109 FT.CREATE rejects RERANK on HNSW vector fields in random access memory (RAM) mode. (MOD-16015)
#Q10200 Disk indexes reject DIALECT 4 and WITHOUTCOUNT. (MOD-15997)
#Q10082 COLLECT requires an @ prefix for field names. (MOD-16233)
#Q11106 JSON HIGHLIGHT/SUMMARIZE fields with multi-value JSONPath are now rejected. (MOD-16530)
#Q10208 Non-cacheable commands are now marked with dont_cache. (MOD-16175)
#Q10699 Numeric command arguments are now strictly parsed. (MOD-17409)
#Q11335 The Search module requires Redis 8.9.241 or newer to load. (MOD-15868)
#P1063 Rejected CMS.INCRBY item updates now leave counters and the total count unchanged instead of saturating counters on overflow. Multi-item commands can still partially succeed.
#P1063 CMS.MERGE now requires matching cell sizes across the destination and all sources.

Modules API

#15517 The modules application programming interface (API) adds the REDISMODULE_NODE_PORT_TLS flag to RM_GetClusterNodeInfo() to retrieve a cluster node's TLS port.
#J1623 Added getWithPath in shared API V9 to evaluate a precompiled JSONPath, allowing callers to compile a path once and reuse it across documents.
#J1666 Added pathParseNoErrorMsg in shared API V10 to compile JSONPaths without a Redis context or an allocated error message. Returns a compiled handle on success or NULL on failure.
#J1653 Made pre-command hooks optional and allowed hook errors to abort JSON commands before execution. Hooks now return RedisResult<()>; shared C API calls continue to ignore hook errors.
#J1640 Extended the Rust Manager interface with clone_value and create_object to support construction of missing paths. Custom implementations must implement these methods and satisfy the new Borrow<Self::V> requirement for owned values.

	Products	Version
	Develop with Redis Clients, data types, programmability, and Redis for AI. RedisVL is listed separately.
	Libraries and tools Client libraries, observability, and cloud integrations. Redis Data Integration is listed separately.
	RedisVL
	Redis Software
	Redis Cloud
	Redis for Kubernetes
	Redis Open Source
	Redis Iris context engine
	Redis Feature Form
	Redis Insight
	Redis Data Integration
	Glossary
	Command reference