Caching

Store expensive results once and serve them from memory or Redis with a type-safe API: cached values are the same model objects as the rest of your code. There is no separate cache client to set up.

Get Started

Type-safe caching

Fetching the same rows on every request can be slow and expensive, but adding a cache usually means a separate client, another serialization format, and hand-written invalidation code. That invalidation code is where stale data and subtle bugs appear.

Serverpod ships a built-in caching system: a local in-memory cache, a priority cache for hot keys, and a distributed cache backed by Redis. They work with your Serverpod models, so cached values are the same type of objects you already pass around.

How it works

1Store a value

Put any serializable model into the cache, optionally with a lifetime.

await session.caches.local.put(
  'product-$id',
  product,
  lifetime: Duration(minutes: 5),
);

2Read it back, or compute on a miss

On a miss, the handler runs, the value is cached, and the same value is returned.

var product = await session.caches.local.get(
  'product-$id',
  CacheMissHandler(
    () async => Product.db.findById(session, id),
    lifetime: Duration(minutes: 5),
  ),
);

3Invalidate when data changes

Remove the entry when the underlying data changes, so readers never see stale values.

await session.caches.local.invalidateKey('product-$id');

Local caches

An in-memory least-recently-used (LRU) cache for general use, plus a priority cache that keeps your hottest keys resident, both scoped per server.

Distributed cache

The same typed API, backed by Redis, gives every server in your cluster a shared cache for session data and expensive computed values.

var product = await session.caches.global.get<Product>('product-$id');

Type-safe values

You cache and read your own model objects, not untyped blobs, so a cached value is checked at compile time just like the objects you store in the database.

Everything included

Why Serverpod

Works with

Redis In-memory caching

Frequently asked questions