CML + OSPF Triangle Lab

Python automation that builds a 3-router OSPF triangle in Cisco Modeling Labs (CML 2.x) end-to-end: create the lab, drop in three IOSv routers, wire them in a triangle, push day-0 configs with OSPF area 0 on every link, and start the lab. One command, reproducible topology.

Script: CML_OSPF_Basic_conf.py

Topology

        R1 (1.1.1.1)
        /          \
 10.0.12.0/30   10.0.13.0/30
      /              \
   R2 (2.2.2.2) --- R3 (3.3.3.3)
          10.0.23.0/30

• Three IOSv routers: R1, R2, R3
• Three /30 point-to-point links in a triangle
• Loopback0 on each router (1.1.1.1, 2.2.2.2, 3.3.3.3) advertised into OSPF
• All links in OSPF area 0 (single area, no summarisation)

Prerequisites

• A reachable CML 2.x controller (tested against 2.7)
• The iosv node definition installed on the controller
• Python 3.9+ and the official CML client:

pip install virl2-client

Configure credentials

Set via environment variables (preferred) or edit the constants near the top of the script:

export CML_URL=https://cml.example.com
export CML_USERNAME=admin
export CML_PASSWORD='changeme'

ssl_verify=False is set in the script for lab use. Flip to True and trust the CA when pointing at a real controller.

Run it

python CML_OSPF_Basic_conf.py

Expected output:

Created lab '3-Router OSPF Triangle' (id=<uuid>)
Nodes and links created. Starting lab...
Lab is running.

Verify once routers boot (~1-2 min):
  R1# show ip ospf neighbor
  R1# show ip route ospf
  R1# ping 3.3.3.3 source 1.1.1.1

The lab appears in the CML UI with the three nodes laid out as a triangle:

What the script actually does

1. Connects to CML using virl2_client.ClientLibrary and waits for the controller to report ready.
2. Creates a lab titled 3-Router OSPF Triangle.
3. Creates three iosv nodes (R1, R2, R3) at fixed coordinates so the layout is always identical.
4. Assigns day-0 configs to each node as a single multi-line string — interfaces, loopback, and router ospf 1 stanza.
5. Pads physical interfaces. IOSv nodes are created with Gi0/0 only; the helper ensure_interfaces() calls create_interface() until each node has three slots (Gi0/0, Gi0/1, Gi0/2).
6. Wires the triangle using lab.create_link() against specific physical_interfaces[n] so the slot numbers line up with what the day-0 configs expect.
7. Starts the lab — nodes boot and load their configs.

Verify OSPF convergence

Once nodes boot (~1–2 minutes), open a console to R1 and check neighbors and routes:

R1# show ip ospf neighbor
Neighbor ID     Pri   State           Dead Time   Address         Interface
2.2.2.2           1   FULL/BDR        00:00:35    10.0.12.2       GigabitEthernet0/1
3.3.3.3           1   FULL/BDR        00:00:38    10.0.13.2       GigabitEthernet0/2

R1# show ip route ospf
     2.0.0.0/32 is subnetted, 1 subnets
O       2.2.2.2 [110/2] via 10.0.12.2, ...
     3.0.0.0/32 is subnetted, 1 subnets
O       3.3.3.3 [110/2] via 10.0.13.2, ...
O    10.0.23.0/30 [110/2] via 10.0.12.2, ...
                         [110/2] via 10.0.13.2, ...

End-to-end reachability across loopbacks confirms the control plane is working:

Extending the lab

Things to bolt on once the basic triangle is healthy:

• More areas — split one link into area 1, add ABR logic and inter-area summarisation (area 1 range ...).
• Authentication — MD5 or SHA on each interface: ip ospf authentication message-digest.
• Timers — dial down hello/dead for sub-second convergence; test with a link flap.
• BFD — enable BFD on the links and have OSPF register for fast failure detection.
• Telemetry — wire each router into the observability stack via IOS-XR MDT (or SNMP for IOSv) and watch neighbor state flap in Grafana as you tear links down.

Troubleshooting

• Node definition 'iosv' not found — install the IOSv image on the CML controller; iosvl2 won't work (layer-2 only).
• Interface index errors — the script assumes create_interface() adds the next sequential slot. If your controller's node definition pre-creates more or fewer interfaces, adjust ensure_interfaces(node, 3).
• Neighbors stuck in EXSTART/EXCHANGE — usually an MTU mismatch. Set matching ip mtu 1500 on both ends or add ip ospf mtu-ignore.
• No OSPF neighbors at all — check the link came up at L1 (show ip interface brief) and that no subinterface or CDP-only VLAN tagging is in the way.