diff --git a/VXLAN/vxlan-ospf/README.md b/VXLAN/vxlan-ospf/README.md
new file mode 100644
index 0000000..85539c4
--- /dev/null
+++ b/VXLAN/vxlan-ospf/README.md
@@ -0,0 +1,23 @@
+# VXLAN Bridging with OSPF
+
+This directory demonstrates what might happen when you run OSPF on VXLAN-enabled VLAN interfaces.
+
+![Lab topology](vxlan-ospf.png)
+
+The directory contains two topologies:
+
+* `topology-broken.yml` runs the same OSPF instance on physical interfaces and VLAN interfaces, resulting in recursive routing
+* `topology-fixed.yml` isolates tenant routing in a VRF resulting in a working network even when S1 and S2 run OSPF across a VXLAN-enabled VLAN
+
+## Changing Device Types
+
+This topology can be used with all network devices supporting VLAN, VXLAN, and OSPF configuration modules, and all virtualization providers supported by *netlab*:
+
+* To change the switch device type, use `-s groups.switch.device=xxx` CLI argument
+* To change the virtualization provider, use `-p` CLI argument.
+
+For example, to start the lab with Nexus 9300v switches using _libvirt_, use:
+
+```
+netlab up -p libvirt -s groups.switches.device=nxos
+```
diff --git a/VXLAN/vxlan-ospf/topology-broken.yml b/VXLAN/vxlan-ospf/topology-broken.yml
new file mode 100644
index 0000000..adadf4e
--- /dev/null
+++ b/VXLAN/vxlan-ospf/topology-broken.yml
@@ -0,0 +1,31 @@
+---
+message: |
+  This topology demonstrates what happens when you run OSPF on
+  "all interfaces" in a VXLAN IRB environment.
+
+provider: clab
+
+vlans:
+  red:
+
+groups:
+  switch:
+    members: [ s1, s2 ]
+    module: [ vxlan, vlan, ospf ]
+    device: eos
+  host:
+    members: [ h1, h2 ]
+    device: linux
+
+nodes: [ s1, s2, h1, h2 ]
+
+links:
+- h1:
+  s1:
+  vlan.access: red
+- h2:
+  s2:
+  vlan.access: red
+- s1:
+  s2:
+  ospf.cost: 2000
diff --git a/VXLAN/vxlan-ospf/topology-fixed.yml b/VXLAN/vxlan-ospf/topology-fixed.yml
new file mode 100644
index 0000000..7e367e9
--- /dev/null
+++ b/VXLAN/vxlan-ospf/topology-fixed.yml
@@ -0,0 +1,35 @@
+---
+message: |
+  This topology demonstrates how to segregate tenant routing from transport
+  (underlay) routing in a VXLAN IRB environment.
+
+provider: clab
+
+vrfs:
+  tenant:
+
+vlans:
+  red:
+    vrf: tenant
+
+groups:
+  switch:
+    members: [ s1, s2 ]
+    module: [ vxlan, vlan, ospf, vrf ]
+    device: eos
+  host:
+    members: [ h1, h2 ]
+    device: linux
+
+nodes: [ s1, s2, h1, h2 ]
+
+links:
+- h1:
+  s1:
+  vlan.access: red
+- h2:
+  s2:
+  vlan.access: red
+- s1:
+  s2:
+  ospf.cost: 2000
diff --git a/VXLAN/vxlan-ospf/vxlan-ospf.png b/VXLAN/vxlan-ospf/vxlan-ospf.png
new file mode 100644
index 0000000..03d9d97
Binary files /dev/null and b/VXLAN/vxlan-ospf/vxlan-ospf.png differ