Keyvalue

Keyvalue, often written as key-value, refers to a simple data model and storage pattern in computing. It stores data as a collection of pairs, where each pair consists of a unique key and an associated value. The key provides fast lookup, while the value can be a string, number, binary blob, or serialized structured data. The model underpins key-value stores, a subset of NoSQL databases, and is widely used for fast direct access to individual records.

Key-value stores expose basic operations such as put (insert or update), get (retrieve by key), and delete. Some implementations also support atomic operations on values, such as incrementing a numeric field or conditional updates. Data can be kept in memory for speed or persisted to disk for durability, and many distributed systems partition and replicate data to scale horizontally and improve fault tolerance.

Characteristics of the key-value approach include a very simple schema, flexible value formats, and fast latency for lookups. However, querying capabilities are usually limited compared with relational databases, and there is typically no built-in support for complex relationships or multi-record transactions. When using distributed stores, trade-offs between latency, consistency, and availability arise (e.g., strong vs. eventual consistency).

Common use cases include caching, session stores, configuration management, feature flags, and storing user preferences or other unstructured data where rapid key-based access is prioritized over complex queries. Notable implementations range from in-memory caches like Memcached to persistent stores like Redis and various cloud-based services that provide key-value interfaces, each with its own consistency and durability characteristics.