Multi-Chassis Link Aggregation
Multi-Chassis Link Aggregation (MLAG) is a solution that allows you to terminate a link aggregation group (sometimes also known as etherchannel) on multiple devices.
It’s often used to implement redundant server connections; it was also popular in the days of layer-2 fabrics built with Spanning Tree Protocol (STP). The latter use case is mostly obsolete in the VXLAN/EVPN world.
What Is Multi-Chassis Link Aggregation?
- Multi-Chassis Link Aggregation (MLAG) Basics
- Multi-Chassis Link Aggregation (MLAG) and Hot Potato Switching
- MLAG and Load Balancing
Technology Deep Dive
- System Overview
- Dynamic MAC Learning
- Layer-2 Flooding
- Layer-3 Forwarding
- Running Routing Protocols over MLAG Links
Using MLAG Clusters with VXLAN and EVPN
- Combining MLAG Clusters with VXLAN Fabric
- Using EVPN/VXLAN with MLAG Clusters
- MLAG Clusters without a Physical Peer Link
Design Guidelines
- BGP AS Numbers on MLAG Members
- Don’t Try to Fake Multi-chassis Link Aggregation (MLAG)
- Stackable Data Center Switches? Do the Math!
- vSphere Does Not Need LAG Bandaids – the Network Might
- Is MLAG an Alternative to Stackable Switches?
- Don't Base Your Design on Vendor Marketing
- Should I Go with VXLAN or MLAG with STP?
- EVPN/VXLAN or Bridged Data Center Fabric?
MLAG Implementations
- Multi-chassis Link Aggregation: Stacking on Steroids
- External Brains Driving an MLAG Cluster
- Intelligent Redundant Framework (IRF) – Stacking as Usual
- Auto-MLAG and Auto-BGP in Cumulus Linux
Failure Scenarios
- Rerouting around LAG Member Failures
- LAG Member Failures in VXLAN Fabrics
- EVPN Rerouting After LAG Member Failures