soundcloud/roshi

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

3180 stars · 155 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Roshi is a distributed CRDT set implementation for timestamped events, built in Go as a stateless layer on top of sharded Redis clusters. It implements a LWW-element-set with limited inline garbage collection, providing fault tolerance via duplicated clusters without consensus protocols.

Use cases

• store timestamped event streams at scale
• build a high-performance index for time-series data
• implement activity feeds like a social stream
• achieve eventual consistency without consensus protocols
• shard datasets across multiple Redis instances with fault tolerance

When to choose

• you need a highly available, partition-tolerant store for timestamped sets in the request path
• you want CRDT semantics to avoid running a consensus protocol
• your dataset is too large for a single Redis instance and needs sharding plus replication

When to avoid

• you need strong consistency or transactions rather than eventual consistency
• your data model needs rich queries beyond ordered set selection by key
• you don't already operate Redis infrastructure and multiple clusters

Facets

library · maturity maintenance

database caching concurrency databases microservices go crdt lww-element-set redis timestamped-events event-stream eventual-consistency real-time linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="soundcloud/roshi")