Using rcfg-sim with rConfig

rcfg-sim was originally built to load-test rConfig, the network configuration management (NCM) platform, and the two pair naturally. This page explains the relationship and how to wire them together.

Two different tools

It’s worth being precise about what each one is:

• rConfig is the product under test — a full network configuration management platform that backs up device configs, tracks changes, runs diffs, and enforces compliance across a real fleet. Its product site is www.rconfig.com, and its user documentation lives at docs.rconfig.com (with the V8 core developer docs at v8coredocs.rconfig.com).
• rcfg-sim (this project) is a test instrument — a simulator that presents a fleet of fake devices so you can drive rConfig (or any SSH-based tool) at scale.

If you’re looking for how to use rConfig itself — installing it, configuring device credentials, scheduling backups, building compliance policies — that’s all in the rConfig documentation, not here. This site is only about the simulator.

The setup

  rConfig            ──SSH backups──▶   rcfg-sim
  (NCM under test)                      (simulated fleet)
      │                                     │
      │ app metrics/logs                    │ /metrics
      └──────────────▶ Prometheus / Grafana ◀┘

1. Generate a fleet that mirrors your target deployment’s shape. Match the device count and pick a size distribution close to your real mix.
2. Expose it on routable IPs so the rConfig host can reach it — see network setup and use routable (not loopback) aliases.
3. Add the devices to rConfig. Use the manifest’s credentials (default admin / admin, enable enable123) and the ip:port pairs from the manifest. The rConfig device-management docs cover bulk import.
4. Run a backup cycle across the whole fleet and watch both sides — rConfig’s own metrics and rcfg-sim’s Prometheus metrics.

What you can learn

• How rConfig’s scheduler behaves when thousands of backups fire in one window
• Throughput and tail latency of a full collection cycle against realistic config sizes
• How the diff and storage layers handle multi-megabyte (xl–6xl) configs
• Resilience to a misbehaving fleet via fault injection — auth failures, mid-transfer drops, slow devices, malformed output

Realistic mixed fleets

For a deployment that includes optical gear, mix in the ciena-6500-tl1 model and run the server with --ssh-auth=driver so Cisco and Ciena devices authenticate the way they really do. This tests rConfig’s multi-vendor handling, not just its Cisco path.

Not an rConfig user?

rcfg-sim speaks plain SSH and emits standard Prometheus metrics — it works with any SSH-based automation, backup, or NCM system. Nothing on this site assumes rConfig beyond this page.