Batching, caching and simplification of Sequelize with facebook/dataloader
dataloader-sequelize is designed to provide per-request caching/batching for sequelize lookups, most likely in a graphql environment
- Should be called after all models and associations are defined
sequelizea sequelize instanceoptions.max=500the maximum number of simultaneous dataloaders to store in memory. The loaders are stored in an LRU cache
import {createContext, EXPECTED_OPTIONS_KEY} from 'dataloader-sequelize';
/* Per request */
const context = createContext(sequelize); // must not be called before all models and associations are defined
await User.findById(2, {[EXPECTED_OPTIONS_KEY]: context});
await User.findById(2, {[EXPECTED_OPTIONS_KEY]: context}); // Cached or batched, depending on timingCommonly you might have some sort of custom findAll requests that isn't going through the dataloader. To reuse the results from a call such as this in later findById calls you need to prime the cache:
import {createContext, EXPECTED_OPTIONS_KEY} from 'dataloader-sequelize';
const context = createContext(sequelize);
const results = await User.findAll({where: {/* super complicated */}});
context.prime(results);
await User.findById(2, {[EXPECTED_OPTIONS_KEY]: context}); // Cached, if was in results
![dependabot[bot] avatar](https://avatars.githubusercontent.com/in/29110?v=4 "dependabot[bot]")
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent