Related Ingenix tool
Document the physical and logical relationships between aggregated links.
In this guide
Why aggregate links?
A Link Aggregation Group combines several physical interfaces into one logical connection. It is useful for switch uplinks, servers and infrastructure where more bandwidth or resilience than one interface provides is required.
How LACP works
LACP peers exchange control information and determine which interfaces can participate in the bundle. Active mode initiates negotiation; passive mode responds. Static aggregation does not provide the same negotiated protection.
| Mode | Behaviour |
|---|---|
| Active | Attempts to negotiate LACP. |
| Passive | Responds to LACP but does not initiate it. |
| Static | Forces aggregation without LACP negotiation. |
Capacity and hashing
A four-port 10GbE LAG provides 40Gb/s of aggregate physical capacity, but a single TCP session will normally use one member. Hash algorithms commonly consider source and destination MAC or IP addresses and transport ports, allowing many independent flows to use different members.
Design considerations
- Keep member speed, MTU and VLAN configuration consistent.
- Use the logical LAG as the configuration boundary where supported.
- Prefer physical path diversity so one failure does not remove every member.
- Confirm both endpoints support the same aggregation method and limits.
- Remember that LAG does not eliminate the need for STP elsewhere in a Layer 2 design.
Troubleshooting
- Confirm expected interfaces are members of the same bundle.
- Check LACP actor and partner state.
- Compare speed, MTU, VLANs and interface settings.
- Look for suspended or detached members.
- Check interface errors and optic health.
- Test several concurrent flows when measuring aggregate throughput.
- Check STP and MAC learning if topology behaviour changes.