Multiget

Multiget is a technique in data storage and retrieval that allows clients to fetch multiple items identified by their keys in a single operation. Instead of issuing separate read requests for each key, a multiget aggregates them, reducing network round trips and improving throughput in latency-sensitive workloads. Multiget is common in in-memory caches, distributed databases, and key-value stores, and it is implemented under various names such as multi-key read, batch get, or MGET.

In Redis, MGET key1 key2 ... returns an array of values in the same order as the requested keys, with null values for missing keys. The operation can be combined with pipeline or transaction features to amortize latency further.

In distributed databases like Cassandra, multiget is achieved through batch reads or multi-key queries. Depending on the data model, such as primary-key based reads or wide-row reads, the system may return results for existing keys while omitting absent ones. Atomicity is not guaranteed across all keys in a batch; consistency may vary with the chosen consistency level.

Similar functionality exists in DynamoDB with BatchGetItem, MongoDB with bulk find, and relational systems via multi-statement or IN clauses, though semantics differ in error handling and limits. Batch operations typically have upper bounds on the number of keys and total payload; partial successes may be reported with per-item status.

Benefits include reduced latency and better throughput for workloads that require many reads. Trade-offs include higher per-operation cost, possible partial failures, and more complex error handling. Multiget does not always guarantee strict atomicity across all keys; developers should consult the specific system’s documentation for semantics and limits.