Pusher-compatible edge
Protocol V1 keeps Channels clients, Laravel Echo, and server SDK expectations intact.
Sockudo is a Rust realtime server for teams that want the familiar Pusher protocol, Protocol V2 durability, horizontal fanout, recovery, push, SDKs, and AI transport in one self-hosted control plane.
sockudo-control
localhost:6001cargo run --release --features fullprotocolVersion: 2idempotency_key: order-createdThe homepage routes operators, backend teams, client developers, and AI product teams into the part of the docs that answers their next production question.
Install Sockudo, choose a runtime, configure apps, and get a local realtime endpoint running.
Open guideUse JavaScript, Swift, Kotlin, Flutter, and .NET clients with Protocol V1 or Protocol V2.
Open guidePlan scaling, metrics, webhooks, rate limits, queues, durable history, and recovery paths.
Open guideLayer agent streams, rollups, push, annotations, and recovery on the same durable primitives.
Open guideSockudo keeps the Protocol V1 contract stable while Protocol V2 unlocks history, recovery, annotations, push delivery, and AI stream primitives on the same fanout core.
Read history and recoveryCompatibility is the on-ramp. The docs also cover the stateful, distributed, and observable pieces that make Sockudo viable when missed events are not acceptable.
Protocol V1 keeps Channels clients, Laravel Echo, and server SDK expectations intact.
Native prefixes, message IDs, serial continuity, filters, deltas, annotations, and rewind.
Redis, Redis Cluster, NATS, Kafka, RabbitMQ, Pulsar, Google Pub/Sub, Iggy, or memory.
Hot replay buffers, durable history, opaque cursors, presence history, and mutable messages.
Device registration, provider status, scheduled notifications, retries, and delivery signals.
Health probes, readiness, Prometheus metrics, webhooks, auth, quotas, and failure surfaces.
Every official SDK path includes installation, configuration, authentication, publish flows, history, and production guidance for teams adopting Sockudo incrementally.
The blog explains the tradeoffs behind the docs: protocol evolution, multi-node operations, SDK boundaries, and the practical shape of ownership.
How Sockudo preserves Pusher compatibility while adding serial continuity, native features, and migration room.
Adapters, duplicate delivery, recovery, observability, and fanout failure modes for production clusters.
Where credentials belong, how client and server responsibilities divide, and which SDK surface to reach for.