mit-pdos/noria

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

5244 stars · 250 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Noria is a streaming data-flow storage backend written in Rust that acts like a database but precomputes and incrementally maintains materialized views of SQL query results. It uses partially-stateful data-flow to keep cached results up-to-date as base tables change, giving read-heavy web applications extremely fast reads and high write throughput.

Use cases

• speed up read-heavy web application backends without manual caching
• replace MySQL or MariaDB with a faster drop-in storage backend
• maintain incrementally-updated materialized views for SQL queries
• cache relational query results like memcached but always consistent
• run Lobsters-like workloads with 5x higher throughput
• dynamically change queries and data-flow at runtime

When to choose

• your web app is bottlenecked on repeated SQL reads and you want automatic, always-fresh caching
• you want materialized views maintained incrementally without manual cache invalidation
• you can adopt a research-grade Rust storage system and run ZooKeeper

When to avoid

• you need a production-hardened database with broad ecosystem support
• your workload is write-heavy or requires full SQL feature coverage
• you cannot operate additional infrastructure like ZooKeeper

Facets

service · maturity maintenance

database caching streaming search-engine databases web-development backend microservices rust self-hosted materialized-views dataflow stream-processing sql mysql-adapter research-system data-engineering linux docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="mit-pdos/noria")