Even after fixing the topology, host bond interfaces still show DOWN due to Docker capability limitations preventing interface enslaving.
Root Cause
ContainerLab's Linux containers don't have NET_ADMIN capability by default, causing Operation not permitted errors when trying to enslave interfaces to a bond.
Solution Applied ✅
Simplified the design - removed bonding entirely and made hosts single-homed:
Hosts now use simple VLAN interfaces directly on eth1 (e.g., eth1.40, eth1.34)
Hosts connect to only one leaf switch (no dual-homing)
MLAG is still configured on leafs but hosts don't use it
New Host Configuration
host1: Connected to leaf1 only, uses eth1.40 for VLAN 40
host2: Connected to leaf3 only, uses eth1.34 for VLAN 34 (VRF gold)
host3: Connected to leaf5 only, uses eth1.40 for VLAN 40
host4: Connected to leaf7 only, uses eth1.78 for VLAN 78 (VRF gold)
This simpler design avoids Docker capability issues while still testing L2 and L3 VXLAN functionality over the EVPN fabric.
Alternative (if dual-homing needed)
To use MLAG with dual-homed hosts, would need to run ContainerLab with privileged containers or add specific capabilities.
## Problem
Even after fixing the topology, host bond interfaces still show DOWN due to Docker capability limitations preventing interface enslaving.
## Root Cause
ContainerLab's Linux containers don't have NET_ADMIN capability by default, causing `Operation not permitted` errors when trying to enslave interfaces to a bond.
## Solution Applied ✅
**Simplified the design** - removed bonding entirely and made hosts single-homed:
- Hosts now use simple VLAN interfaces directly on eth1 (e.g., eth1.40, eth1.34)
- Hosts connect to only one leaf switch (no dual-homing)
- MLAG is still configured on leafs but hosts don't use it
### New Host Configuration
- **host1**: Connected to leaf1 only, uses eth1.40 for VLAN 40
- **host2**: Connected to leaf3 only, uses eth1.34 for VLAN 34 (VRF gold)
- **host3**: Connected to leaf5 only, uses eth1.40 for VLAN 40
- **host4**: Connected to leaf7 only, uses eth1.78 for VLAN 78 (VRF gold)
This simpler design avoids Docker capability issues while still testing L2 and L3 VXLAN functionality over the EVPN fabric.
## Alternative (if dual-homing needed)
To use MLAG with dual-homed hosts, would need to run ContainerLab with privileged containers or add specific capabilities.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Problem
Even after fixing the topology, host bond interfaces still show DOWN due to Docker capability limitations preventing interface enslaving.
Root Cause
ContainerLab's Linux containers don't have NET_ADMIN capability by default, causing
Operation not permittederrors when trying to enslave interfaces to a bond.Solution Applied ✅
Simplified the design - removed bonding entirely and made hosts single-homed:
New Host Configuration
This simpler design avoids Docker capability issues while still testing L2 and L3 VXLAN functionality over the EVPN fabric.
Alternative (if dual-homing needed)
To use MLAG with dual-homed hosts, would need to run ContainerLab with privileged containers or add specific capabilities.