Design a highly scalable, fault-tolerant distributed caching system that stores key-value pairs in memory and provides low-latency access for read-heavy applications.
Functional Requirements
Non-Functional Requirements
Extra
Redis can run as a single node, with a high availability (HA) replica, or as a cluster. When operating as a cluster, Redis clients cache a set of "hash slots" which map keys to a specific node. This way clients can directly connect to the node which contains the data they are requesting.
Redis is really, really fast. Redis can handle O(100k) writes per second and read latency is often in the microsecond range. This scale makes some anti-patterns for other database systems actually feasible with Redis. As an example, firing off 100 SQL requests to generate a list of items with a SQL database is a terrible idea, you're better off writing a SQL query which returns all the data you need in one request. On the other hand, the overhead for doing the same with Redis is rather low - while it'd be great to avoid it if you can, it's doable.
When using Redis as a cache, you'll often employ a time to live (TTL) on each key. Redis guarantees you'll never read the value of a key after the TTL has expired and the TTL is used to decide which items to evict from the server - keeping the cache size manageable even in the case where you're trying to cache more items than memory allows.