REFERENCE ARCHITECTURE
Real-time API & notifications
WebSocket/long-poll API on Compute, Beacon for pub/sub fan-out, Lane for durable notification delivery, Redis state in Datastore.
IntermediatePlan: Business+Resilience: multi-zona
Diagram
Use cases
- Live ops dashboard
- In-app notifications
- Collaboration presence
Components
| Layer | Role | Product |
|---|---|---|
| Edge | TLS + routing | Gateway |
| Realtime | WebSocket API | Compute |
| Pub/sub | Event fan-out | Beacon |
| Queue | Offline delivery | Lane |
| State | Sessions & presence | Datastore (Redis) |
| Scale | Connection bursts | Autoscale |
| Operations | Connection metrics | Insight |
Products
Compute
View product docs →Beacon
View product docs →Datastore
View product docs →Gateway
View product docs →Autoscale
View product docs →Insight
View product docs →When to use
- Live dashboards, chat, collaborative editing presence.
- Mobile push notification backends.
- IoT telemetry fan-out to thousands of subscribers.
When to avoid
- Batch-only reporting — use data pipeline pattern.
- Strict ordering across all users globally — Conduit may fit better.
Design notes
Beacon for fan-out
Publish domain events to Beacon subjects; edge Compute instances subscribe and push to connected clients.
Lane for reliability
When a user is offline, enqueue notifications in Lane; worker Compute retries with exponential backoff.
Blueprint
Use this JSON as a starting point when creating a project via the Tesserra API or console. Replace image URLs, domains, and resource references with your values.
json
{
"nome": "Realtime Platform",
"recursos": [
{
"tipo": "compute",
"nome": "realtime-api",
"config": {
"subtipo": "api",
"porta": 8080,
"replicas_min": 2,
"replicas_max": 16
}
},
{
"tipo": "compute",
"nome": "notify-worker",
"config": {
"subtipo": "worker",
"replicas_min": 1,
"replicas_max": 8
}
},
{
"tipo": "beacon",
"nome": "events",
"config": {
"jetstream": true
}
},
{
"tipo": "lane",
"nome": "offline-notify",
"config": {
"duravel": true
}
},
{
"tipo": "datastore",
"nome": "presence",
"config": {
"engine": "redis",
"tamanho_gb": 10
}
},
{
"tipo": "gateway",
"nome": "realtime-dns",
"config": {
"dominio": "live.acme.io",
"tls": true
}
},
{
"tipo": "autoscale",
"nome": "ws-scale",
"config": {
"alvo_recurso_id": "realtime-api",
"metrica": "connections",
"limiar": 70
}
},
{
"tipo": "insight",
"nome": "live-obs",
"config": {
"retencao_dias": 14,
"alertas": true
}
}
]
}Related
Event-driven integration
Asynchronous workflows with Conduit (Kafka), Lane (RabbitMQ), and Beacon (NATS) — decouple producers from consumers and scale workers independently.
AI inference workload
Serve models via Compute (self-hosted) or Tesserra managed agents — cache with Datastore Redis, audit in Postgres, datasets in Archive.
B2B SaaS multi-tenant platform
Tenant isolation at the organization layer, Fabric private networking, per-tenant Datastore schemas, Gateway custom domains, Sentinel and Autoscale for enterprise SLAs.