A Redis module for Azure Managed Redis that provides secondary indexes plus full-text and vector search. It can only be turned on when the cache is created and needs Enterprise clustering with the NoEviction policy.
Read more: Microsoft Learn
In the Ultra Transcenders books
Each book explains RediSearch in context, with comparison tables and the common traps.
Terms in this definition
- Azure Managed Redis
Successor to Azure Cache for Redis: a managed, Redis Enterprise-based in-memory store, with Memory Optimized, Balanced, Compute Optimized and Flash Optimized tiers.
- Vector search
Searching embeddings for similar meaning, so paraphrases are found via nearest-neighbour algorithms like HNSW; exact identifiers, however, may be missed.
- Enterprise
Any group of organisations with shared goals. Examples range from an entire company or one of its divisions to a government department, or several organisations working as partners or along a supply chain.
- NoEviction
A Redis eviction policy under which writes are refused, rather than keys removed, once memory is exhausted. Azure Managed Redis needs it, together with the Enterprise clustering policy, for RediSearch.
- Exclusions
Scopes taken out of a policy assignment; they can only narrow its coverage, never widen it.
Related terms
- Clustering policy
A choice made when creating Azure Managed Redis (OSS, Enterprise or Non-clustered) that controls how clients see its shards. RediSearch needs Enterprise, and an OSS or Enterprise choice is permanent.
- Eviction policy
The Redis setting that decides which keys get dropped once memory runs out. Azure Managed Redis uses volatile-lru by default, which removes only keys carrying a TTL, while RediSearch needs NoEviction.
- FLAT
A RediSearch vector index that compares every vector exactly, by brute force. It fits smaller data sets or cases needing complete results, whereas HNSW is approximate.
- HNSW
Graph-structured index for approximate vector search in RediSearch and pgvector. Compared with IVFFlat, recall versus speed is better, though building takes longer and memory use is higher.
- KNN
A vector query returning the K vectors closest to a given one, written in RediSearch as [KNN K @field \$vec].
- OSS clustering policy
Exposes the Redis Cluster API in Azure Managed Redis, so clients reach shards directly and get maximum throughput. RediSearch is unavailable with it, and it is fixed at creation time.
- RedisJSON
Lets Azure Managed Redis (any tier) hold nested JSON documents and update parts of them, and RediSearch can index those documents. The module must be chosen when the cache is created.
- Semantic caching
A pattern, in an application or gateway, that returns a previous LLM answer when a new prompt means much the same thing, using RediSearch vector search. Azure Managed Redis doesn't offer it as a feature in its own right.