From Git repository
to live URL in 1 Go.

Push code, isolate heavy builds inside on-demand AWS ECS containers, stream live logs via Apache Kafka & Redis, and serve assets through dynamic Nginx edge proxies.

deploy1t-engine // pipeline-telemetry

Real-time build log ingestion • Topic: container-logsLIVE

01[ECS Task]Provisioning Fargate runner (0.5 vCPU, 1GB RAM)...

02[Git Engine]Cloning repository: main branch verified

03[Kafka Stream]Topic 'container-logs' ingested payload (offset #1042)

04[Build Server]Running: npm install && npm run build (12.4s)

05[ClickHouse]Persisted 248 log lines to MergeTree cold-store

06[AWS S3]Uploaded bundle to __outputs/my-portfolio/

07[Nginx Proxy]Routed wildcard: http://my-portfolio.localhost:8000

08[Status]Container exited (0 idle cost) • Status: READY

THE 3-SERVICE ARCHITECTURE

Engineered for absolute isolation & speed

Ephemeral Build Containers

Decoupled from API orchestration. Builds compile inside isolated AWS ECS Fargate tasks with dynamic environment injection, then terminate immediately.

Memory Isolation100% Sandbox

Dual-Path Log Pipeline

Apache Kafka buffers high-throughput log streams. Consumer bridges hot-path logs to Redis Pub/Sub for SSE streaming, and micro-batches cold logs into ClickHouse.

Log DeliveryHot SSE + Cold Columnar

Dynamic Wildcard Edge

Nginx regex captures tenant subdomains dynamically (`*.localhost:8000`), routing to S3 outputs with SPA client-side fallback and zero config reloads.

DNS Reload Required0 ms (Zero-Reload)

~3 min

Avg Build & Deploy Time

<50ms

SSE Live Log Latency

100%

Containerized Isolation

MergeTree

ClickHouse Historical Retention

Ready to deploy your next project?