marten-querying

To query Marten documents and projections efficiently, you can use standard LINQ for everyday cases, compiled queries for hot paths, batched queries for multiple round trips, and raw SQL for complex logic. This Skill streamlines the entire query process for read-side endpoints.

The best way to optimize frequently called Marten queries is by using compiled queries. Compiled queries cache the query plan for hot paths, significantly improving performance over standard LINQ for repetitive data retrieval operations.

Yes, you can execute multiple independent Marten queries in a single round trip using batched queries. This approach reduces network latency and improves efficiency when you need to retrieve various documents or projections simultaneously.

You can stream JSONB data directly from Marten to an HTTP response for high-performance endpoints. This Skill supports accessing raw JSONB data, allowing you to forward it directly to caches or interface with non-.NET code without deserialization overhead.

You should use raw SQL instead of LINQ to query a Marten event store when you need full control over complex query logic that standard LINQ cannot express. This approach is suitable for complex analytics and data retrieval scenarios requiring database-specific features.

Yes, this Skill supports reusable query plans that can run from multiple call sites with a single source of truth. This ensures consistency across your Marten-based application when executing identical data retrieval operations from different endpoints.