According to the (hypothetical) product team:
The Orchestrator870OVA appliance is built on a modular microservices architecture, containerized internally but presented as a single VM. From a high level, it contains:
| Component | Technology | Function | |-----------|------------|----------| | API Gateway | Traefik + OAuth2 Proxy | Authentication, rate limiting, routing | | Scheduler | Custom Go scheduler | Trigger workflows (time, event, or API) | | Executor | Celery + Redis | Task queue and distributed execution | | Worker Pool | Python multiprocessing | Local execution of tasks (or remote via SSH) | | Metadata DB | PostgreSQL 15 | Workflow definitions, run history, variables | | Artifact Store | MinIO | Logs, cached data, task outputs | | Dashboard | React + WebSockets | Real-time DAG visualization, logs, retries | | Monitoring | Prometheus + Grafana | Metrics (workflow duration, task failures, queue length) | orchestrator870ova
All components communicate via gRPC for low-latency internal calls, with fallback to REST.
![Conceptual architecture of Orchestrator870OVA – centralized scheduler, executor queue, worker pool, and external connectors.] According to the (hypothetical) product team:
Orchestrator870OVA ships with 30+ native connectors, including:
| Category | Examples | |----------|----------| | Cloud | AWS (Lambda, S3, EC2), Azure (Blob, Functions), GCP (BigQuery, Cloud Run) | | Databases | PostgreSQL, MySQL, Snowflake, Redshift, MongoDB | | Messaging | Kafka, RabbitMQ, SQS, NATS | | CI/CD | Jenkins, GitLab CI, GitHub Actions (webhook triggers) | | Monitoring | Datadog, New Relic, Prometheus (remote write) | The Orchestrator870OVA appliance is built on a modular
Custom connectors can be written in Python or Node.js and added as a mounted volume (no rebuild of appliance required).