NXOS port-channel bandwidth mismatch when interface is down

[dhalperi]

Problem

Batfish and NXOS show different bandwidth values for down port-channel interfaces, causing validation discrepancies.

Technical Details

When a port-channel interface is administratively down or has no active members, NXOS assigns a default bandwidth value (typically 100000000 bps) while Batfish reports 0.0 bps bandwidth.

Observed In

NXOS lab configurations with down port-channel interfaces

Symptoms

• Interface validation tests fail with bandwidth mismatches like:

  'port-channel54': {'bandwidth': 'Batfish: 0.0, NXOS: 100000000'}
  

• Only affects port-channel interfaces that are not active
• Current workaround: NxosValidator skips bandwidth validation for inactive port-channel interfaces (line 221-225 in NxosValidator.py)

Configuration Patterns

Port-channel interfaces without active member links or in administratively down state

Next Steps

Determine if Batfish should model default bandwidth for down port-channel interfaces to match NXOS behavior, or if the current filtering approach is preferred for validation purposes.