ORelay

Streaming made simple

Example Guide

Streaming, as you like it

Create streams the way you want them: WebSockets, HTTP requests, Kafka, HTTP 2, HTTP 3, WebRTC, Spark, and many more.

Inbound, outbound

Wire your components together and forget if they are services or clients. You can even use Cron Expressions to make 2 APIs stream to each other - without a background ticker publishing to ORelay.

Synchronous or asynchronous?

Synchronous requests are so convenient! Why not use ORelay to keep that simplicity?

Scalability

ORelay is built with scalability in mind. You won't need to explicitly declare partitions or shards. Just watch ORelay rebalance as replicas are added to the cluster.

Request Batching

If a single request takes 20ms, but a batch of 100 takes 30ms, then you would prefer to batch your requests, right? Gotcha fam.

Rust all the way

Fans of Ferris and followers of Crabulon will agree this is exciting. Rust brings performance and memory-safety to ORelay, and underpins what makes this software so fast.

Near-zero integration

Integration is scary. That is why ORelay tries to keep things simple. You should be able to simply point your endpoints at the service and start streaming.

Stability

ORelay is rigorously tested, the same way databases are. Randomised tests explore any possible edge cases, while also validating correctness of data.