effect-batching

To solve the N+1 query problem in Effect TypeScript, use RequestResolver to batch concurrent Effect.request calls together. This Skill collects concurrent lookups and resolves them in a single execution, deduplicating identical requests inside the batch window to prevent redundant fetches.

Request batching and deduplication works by collecting concurrent Effect.request calls into a single resolver execution window. Identical requests made within the configured batch delay are automatically deduplicated, ensuring only one execution occurs per unique request.

You configure batch size and delay for resolvers in Effect by tuning the RequestResolver settings. You can adjust the batch window delay and maximum batch size to control how many concurrent requests are collected together before execution.

Yes, effect-batching supports SqlResolver for schema-validated batched SQL execution. It provides ordered, grouped, findById, and void resolver variants with transaction awareness to integrate batching directly into your SQL data access layer.

Yes, you can add tracing and observability to batched resolver execution by instrumenting resolvers with spans. This allows you to trace batch execution details, monitor cache behavior, and observe batch delay and size performance metrics.

If a resolver does not complete every entry in a batch, the execution is considered incomplete. The Skill requires that every resolver entry is properly completed, returning typed errors for misses such as when a requested ID is not found in the data source.