How MpegFlow fits in your stack.

Object storage for mezzanine assets, encoded outputs, and archive footage. The data plane for every video pipeline.

• AWS S3
  MpegFlow with AWS S3: video transcoding architecture
  How MpegFlow integrates with AWS S3 — presigned URL pattern, multi-region replication, lifecycle policies, and the IAM-zero strict-broker pattern for production video pipelines.
• Cloudflare R2
  MpegFlow with Cloudflare R2: zero-egress video storage
  How MpegFlow integrates with Cloudflare R2 — the S3-compatible API, zero-egress economics for CDN delivery, and the multi-cloud benefits over AWS S3 lock-in.
• Google Cloud Storage
  MpegFlow with Google Cloud Storage (GCS)
  How MpegFlow integrates with Google Cloud Storage — interoperability via S3 API, multi-region buckets, and dual-region for active-active video workloads.
• MinIO
  MpegFlow with MinIO: self-hosted S3-compatible video storage
  How MpegFlow integrates with MinIO for self-hosted, on-prem, or air-gapped video workloads. S3-compatible API, erasure coding, and the sovereign-cloud deployment shape.

Where the encoder pool runs — Kubernetes is the production answer; bare metal and VMs work for small deployments.

• Kubernetes
  MpegFlow on Kubernetes: production deployment topology
  How MpegFlow runs on Kubernetes — Helm chart deployment, KEDA autoscaling, the operator pattern for multi-tenant pools, and the network policy enforcement that makes strict-broker security work.

Job state, audit log, and queue infrastructure. PostgreSQL for state, Redis for queues + ephemeral state.

• PostgreSQL
  MpegFlow with PostgreSQL: job state, audit log, multi-tenant tables
  How MpegFlow uses PostgreSQL for job state, the append-only audit log, multi-tenant table layout, and the HA patterns that survive production traffic.
• Redis
  MpegFlow with Redis: queues, distributed locks, real-time state
  How MpegFlow uses Redis for queues (job dispatch), distributed locks (operator coordination), and real-time state (worker heartbeats). The HA patterns that survive failover events.

Metrics, traces, and logs. The instrumentation that lets you debug production issues at 3am.

• Datadog
  MpegFlow with Datadog: metrics, APM, log aggregation
  How MpegFlow integrates with Datadog — OpenMetrics scraping, distributed tracing via OTLP, log aggregation, and the dashboards that matter for video pipeline ops.
• Prometheus + Grafana
  MpegFlow with Prometheus + Grafana: open-source observability
  How MpegFlow integrates with Prometheus + Grafana — the open-source observability stack. Native OpenMetrics, recording rules, the dashboards that work, and when this beats Datadog.

Infrastructure-as-code and chart-based deployment. The DevOps layer that makes upgrades reviewable.

• Terraform
  MpegFlow with Terraform: infrastructure as code
  How to deploy MpegFlow via Terraform — the providers, the module patterns, and the GitOps integration that makes infrastructure changes reviewable.
• Helm
  MpegFlow with Helm: chart structure and deployment
  The MpegFlow Helm chart — what it deploys, how to configure it, and the values.yaml shape that production deployments override. From hello-world to multi-tenant.