Chapter 16: EtherChannel
The introduction explains why EtherChannel groups redundant physical links so they can remain active without creating Layer 2 loops.
After this chapter you will be able to
- EtherChannel logical link grouping. Explain how EtherChannel combines physical links into one logical link and why this lets the links remain active without causing Layer 2 loops.
- EtherChannel and the STP bottleneck. Explain why adding separate links between two switches does not relieve congestion when STP blocks the redundant links, and how EtherChannel allows the links to remain active as one logical link.
- Frame forwarding across member links. Describe how frames use the member links of an EtherChannel and why a broadcast frame does not loop among those links.
- Dynamic and static EtherChannel. Distinguish dynamic EtherChannel, in which switches negotiate to form an EtherChannel, from static EtherChannel, in which specified ports form one without negotiation with the neighboring device.
- Dynamic negotiation and protocol choice. Explain why dynamic EtherChannel is recommended over static EtherChannel and distinguish when to use PAgP or LACP.
- PAgP and LACP mode combinations. Describe which PAgP and LACP mode pairs can form an EtherChannel and why two non-initiating modes cannot.
- Channel groups and logical interfaces. Describe how assigning member ports to a channel group creates a logical port-channel interface and what the group number means.
- EtherChannel status and neighbor checks. Explain how the section’s show commands and status flags indicate whether an EtherChannel has formed and which neighbor mode is in use.
- Port-channel configuration inheritance. Explain how port-channel and member-port settings are inherited and why changes to an active EtherChannel should generally be made on the port-channel interface.
- Static EtherChannel formation. Explain that a static EtherChannel requires on mode at both ends and forms without negotiation, regardless of the neighboring device’s state.
- Static EtherChannel risk. Justify the recommendation to use PAgP or LACP instead of a static EtherChannel by describing their configuration checks and the possible consequence of a misconfiguration.
- Matching EtherChannel port settings. Explain which kinds of settings must match on EtherChannel ports and what happens when a member port’s operational settings differ.
- Per-flow physical link selection. Trace how a frame selected for an EtherChannel is assigned to a physical link, and explain why messages in the same flow use that link.
- Total capacity versus flow speed. Distinguish an EtherChannel’s total bandwidth from the maximum speed of an individual communication, and explain the possible effect on congestion.
- Layer 3 EtherChannel purpose. Explain how a Layer 3 EtherChannel differs from a Layer 2 EtherChannel and why it can still be useful even though routed ports are not disabled by STP to prevent Layer 2 loops.
- Routed EtherChannel configuration and status. Describe the additions needed to configure a Layer 3 EtherChannel on a multilayer switch and how the section identifies one in status output.
What it builds on
- The need for STP (chapter 14) · The need to prevent Layer 2 loops in networks with redundant connections.
- Frame flooding and forwarding (chapter 6) · Switches flood frames out eligible ports.
- The STP algorithm (chapter 14) · STP selects which paths can forward without forming loops.
- DTP negotiation (chapter 13) · The familiar distinction between negotiated and manually configured trunk ports supports the section’s analogy.
- Configuring trunk ports (chapter 12) · Configuring switch ports as trunks.
Section by section
Introduction
The introduction explains why EtherChannel groups redundant physical links so they can remain active without creating Layer 2 loops.
- STP redundancy tradeoff: Explain why STP prevents loops but leaves redundant links unused for forwarding until an active link fails.
Why it matters: This limitation explains why EtherChannel is useful. - EtherChannel logical link grouping: Explain how EtherChannel combines physical links into one logical link and why this lets the links remain active without causing Layer 2 loops.
Why it matters: Grouping links allows better use of available bandwidth while retaining redundancy.
Key terms: STP, redundant links, Layer 2 loops, EtherChannel, physical links, logical link
16.1 How EtherChannel works
This section explains how EtherChannel combines physical links to relieve a congestion bottleneck while still allowing STP to prevent Layer 2 loops.
- EtherChannel and the STP bottleneck: Explain why adding separate links between two switches does not relieve congestion when STP blocks the redundant links, and how EtherChannel allows the links to remain active as one logical link.
Why it matters: It explains why simply plugging in more links may not fix an inter-switch bottleneck. - Frame forwarding across member links: Describe how frames use the member links of an EtherChannel and why a broadcast frame does not loop among those links.
Why it matters: It shows how multiple active member links can carry frames without creating a Layer 2 loop. - STP treatment of port channels: Explain why STP can still block an EtherChannel in a larger LAN and how an EtherChannel appears in STP output.
Why it matters: Bundling links does not eliminate STP's role in preventing loops across a larger network.
Key terms: Spanning Tree Protocol, EtherChannel, physical links, logical link, bandwidth, member ports, load balancing, load distribution, BUM frames, STP
16.2 EtherChannel configuration
This section distinguishes dynamic from static EtherChannel configuration on Cisco switches.
- Dynamic and static EtherChannel: Distinguish dynamic EtherChannel, in which switches negotiate to form an EtherChannel, from static EtherChannel, in which specified ports form one without negotiation with the neighboring device.
Why it matters: The distinction tells you whether forming the EtherChannel depends on negotiation with the neighboring switch.
Key terms: dynamic EtherChannel, static EtherChannel, negotiation
16.2.1 Dynamic EtherChannel
This section explains how PAgP and LACP negotiate dynamic EtherChannels, how to configure and verify them, and how to manage their logical port-channel interfaces.
- Dynamic negotiation and protocol choice: Explain why dynamic EtherChannel is recommended over static EtherChannel and distinguish when to use PAgP or LACP.
Why it matters: These distinctions help you choose a compatible protocol and avoid unsafe EtherChannel configurations. - PAgP and LACP mode combinations: Describe which PAgP and LACP mode pairs can form an EtherChannel and why two non-initiating modes cannot.
Why it matters: Mode behavior lets you predict whether two configured switches will actually form a bundle. - Channel groups and logical interfaces: Describe how assigning member ports to a channel group creates a logical port-channel interface and what the group number means.
Why it matters: The logical interface is the object you configure and inspect once the physical links are bundled. - EtherChannel status and neighbor checks: Explain how the section’s show commands and status flags indicate whether an EtherChannel has formed and which neighbor mode is in use.
Why it matters: Verification distinguishes ports merely configured for a channel from ports successfully bundled into one. - Port-channel configuration inheritance: Explain how port-channel and member-port settings are inherited and why changes to an active EtherChannel should generally be made on the port-channel interface.
Why it matters: Making active-bundle changes in one place helps preserve consistent settings and avoid a brief outage. - Protocol link capacity limits: State the different member-link limits for PAgP and LACP, including how many LACP links can be active at once.
Why it matters: These limits prevent confusing the number of links LACP supports with the number it can use simultaneously.
Key terms: dynamic EtherChannel, static EtherChannel, PAgP, LACP, desirable, auto, active, passive, channel-group, group-number
16.2.2 Static EtherChannel
This section explains how static EtherChannels form without negotiation, why that creates risk, and why they cannot pair with PAgP or LACP.
- Static EtherChannel formation: Explain that a static EtherChannel requires on mode at both ends and forms without negotiation, regardless of the neighboring device’s state.
Why it matters: Static configuration does not confirm that the device at the other end is configured appropriately. - Static EtherChannel risk: Justify the recommendation to use PAgP or LACP instead of a static EtherChannel by describing their configuration checks and the possible consequence of a misconfiguration.
Why it matters: The missing configuration check can turn a bundling mistake into a serious LAN problem. - Static and dynamic compatibility: State whether a static EtherChannel can work when the other switch uses PAgP or LACP.
Why it matters: The two ends must use compatible EtherChannel modes for the connection to operate.
Key terms: static EtherChannel, on mode, PAgP, LACP, Layer 2 loops
16.2.3 Physical port configurations
This subsection explains why EtherChannel member ports need matching operational settings and what happens when one port’s settings differ.
- Matching EtherChannel port settings: Explain which kinds of settings must match on EtherChannel ports and what happens when a member port’s operational settings differ.
Why it matters: Matching operational settings keeps a port active in the EtherChannel and able to forward frames.
Key terms: EtherChannel, native VLAN, suspended
16.3 EtherChannel load balancing
This section explains how EtherChannel selects a physical link for each traffic flow, how that selection can be checked or changed, and why added links increase total capacity without increasing a single flow’s maximum link speed.
- Per-flow physical link selection: Trace how a frame selected for an EtherChannel is assigned to a physical link, and explain why messages in the same flow use that link.
Why it matters: It explains how one EtherChannel can carry multiple communications across its member links. - Load-balancing parameter choices: Describe how the switch’s load-balancing parameters can be checked or changed, including what happens when a frame does not contain an IP packet.
Why it matters: The chosen parameters affect how traffic is distributed, and the displayed setting reveals what the switch uses. - Total capacity versus flow speed: Distinguish an EtherChannel’s total bandwidth from the maximum speed of an individual communication, and explain the possible effect on congestion.
Why it matters: It prevents mistaking higher aggregate capacity for a higher maximum rate for one communication.
Key terms: frame-forwarding logic, EtherChannel load-balancing logic, flow, physical link, port-channel load-balance, show etherchannel load-balance, src-dst-ip, Non-IP: Source XOR Destination MAC address, bandwidth, speed
16.4 Layer 3 EtherChannel
This section explains what Layer 3 EtherChannels are, why they are useful, and how to configure and recognize them on multilayer switches.
- Layer 3 EtherChannel purpose: Explain how a Layer 3 EtherChannel differs from a Layer 2 EtherChannel and why it can still be useful even though routed ports are not disabled by STP to prevent Layer 2 loops.
Why it matters: The distinction explains when bundling routed links is useful despite the absence of a Layer 2 loop problem. - Routed EtherChannel configuration and status: Describe the additions needed to configure a Layer 3 EtherChannel on a multilayer switch and how the section identifies one in status output.
Why it matters: Putting the IP address on the logical interface and recognizing its routed status are central to configuring and checking the bundle.
Key terms: Layer 3 EtherChannel, routed ports, Layer 2 EtherChannel, STP, no switchport, port channel interface, PAgP, LACP, on mode, RU
16.5 Summary
Concepts explained in this chapter
- etherchannelA Layer 3 EtherChannel is a logical link aggregation of routed ports that forwards packets at the network layer.
- port-channel interfaceA port-channel interface is a logical Cisco IOS interface representing an aggregated group of physical links.
- dynamic etherchannelA dynamic EtherChannel is an EtherChannel configuration in which neighboring switches negotiate with each other using a protocol to form the logical link.
- PAgPPAgP is Cisco's proprietary protocol for negotiating the formation of an EtherChannel between switch ports.
- channel-group commandThe channel-group command assigns a switch port to an EtherChannel and specifies its negotiation or static mode.
- lacpLACP is a link-aggregation control protocol that negotiates and maintains bundled network links between compatible devices.
- show etherchannel summaryshow etherchannel summary is a Cisco IOS command that displays the status and member-port information of EtherChannels.
- information flowA flow is a communication between two network nodes whose traffic can be assigned consistently to a physical link.
- port-channel load-balanceport-channel load-balance is a Cisco IOS global configuration command that configures the traffic fields used to select EtherChannel member links.
- network bandwidthBandwidth is the total data-carrying capacity of a network connection, measured as the number of bits transferable per second.
- static etherchannelA static EtherChannel is an EtherChannel configuration that forms a logical link from specified ports without negotiating with the neighboring device.
- native vlanA native VLAN is the VLAN assigned to untagged traffic received on a trunk port and the VLAN whose forwarded traffic is sent untagged on that trunk.
- active modeActive mode is an LACP setting that actively attempts to negotiate an EtherChannel with a neighboring switch.
- auto modeAuto mode is a PAgP setting that passively responds to negotiation messages from a desirable-mode neighbor.
- desirable modeDesirable mode is a PAgP setting that actively attempts to negotiate an EtherChannel with a neighboring switch.
- passive modePassive mode is an LACP setting that responds to negotiation messages from an active-mode neighbor without initiating negotiation.
- spanning tree protocolThe spanning tree algorithm is the procedure for electing a root bridge and selecting root and designated ports to create a loop-free spanning-tree topology.
- allowed vlan listAn allowed VLAN is a VLAN permitted to traverse a trunk interface.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.