Redis
In-memory store for caches, queues, and rate limits.
- made by
- Redis
- source
- Official Redis site
Magic Ship is one shop in Vancouver, BC, working remotely with clients worldwide. We are not a partner, reseller, or certified vendor of Redis - we just build with this.
What Redis is
Redis is an in-memory data store used for caching, queues, rate limiting, session state, and pub/sub, with optional persistence to disk. A licence change in 2024 prompted the Valkey fork, which is API-compatible for the common cases and maintained under the Linux Foundation.
How we use it
We put it in front of expensive work: caching embeddings and model responses keyed by a hash of the input, rate limiting per tenant, and holding job state for background workers. Idempotency keys live here so a retried webhook does not run the same side effect twice. Anything that must survive a restart goes into Postgres instead.
Where it is the wrong choice
It is priced by memory, so caching large payloads gets expensive quickly, and default persistence settings can lose the most recent writes in a crash. For a durable queue where every message matters, Postgres or a real broker is the safer choice.
Service lines it turns up in
Related tools
More in Data and backend
Other tools in the same service lines
Building something on Redis?
Send the problem rather than a job spec. You get an answer on scope, on fit, and on whetherRedis is even the right call for it.
Start a projectRedis and Redis are trademarks of their respective owners, used here to say what we work with.