Hello! Welcome to the final lesson in our module on Advanced LAN Switching.
In the past few lessons, we've learned how to configure Layer 2 EtherChannels using LACP and PAgP, and how to optimize their performance by tuning load-balancing methods. Now, we'll address the crucial final step: what to do when things don't work as expected.
This lesson focuses on the learning outcome: Troubleshoot EtherChannel misconfigurations and inconsistent settings. We will explore the common pitfalls in EtherChannel configuration, learn how to interpret the diagnostic signs, and build a systematic workflow for identifying and resolving these issues. This is a vital skill, as a misconfigured EtherChannel can not only fail to provide the intended benefits but can also actively harm your network by creating switching loops.
1. The Primary Diagnostic Tool: show etherchannel summary
Your first and most important command when troubleshooting an EtherChannel is show etherchannel summary. The output provides a concise overview of all Port-channels on the switch and the status of their member ports. The key to using this command effectively is understanding the status flags.
To start, let's get familiar with these flags and what they signify.
EtherChannel // CCNP ENCOR 350-401 Complete Course
This video from Jeremy's IT Lab provides an excellent tour of the show etherchannel summary command. It clearly explains the meaning of the most common flags you will encounter.
Watch from 26:56 to 29:58. Focus on understanding the difference between flags for the Port-channel interface itself (e.g., S, U, D) and flags for the member ports (e.g., P, I, s, w).
Here is a quick reference table for the flags discussed in the video:
| Flag | Meaning | Location | Good or Bad? |
|---|---|---|---|
| S | Layer 2 | Port-channel | Good (Indicates a switch port EtherChannel) |
| U | In use | Port-channel | Good (Indicates the channel is operational) |
| D | Down | Port-channel | Bad (Indicates the channel is not operational) |
| P | In port-channel | Member Port | Good (Indicates the port is bundled) |
| I | Stand-alone | Member Port | Bad (Port is not bundled with its peer) |
| s | Suspended | Member Port | Bad (Port is incompatible with the bundle) |
| w | Waiting | Member Port | Informational (LACP/PAgP negotiation is ongoing) |
With this foundation, let's explore the common issues that cause these "bad" flags to appear.
2. Common Misconfigurations and Their Symptoms
Most EtherChannel problems fall into one of two categories: mismatched parameters on member ports, or mismatched protocol settings between switches.
2.1. Inconsistent Port Parameters (The s Flag)
For ports to be bundled together in an EtherChannel, they must have compatible configurations. Key parameters that must match across all member links include:
- Speed and Duplex
- VLAN configuration (access VLAN or trunking mode with native/allowed VLANs)
- Layer 2 vs. Layer 3 mode
If you configure a port with settings that are inconsistent with the other members of the bundle, the switch will prevent it from joining the channel by placing it in a suspended state.
EtherChannel // CCNP ENCOR 350-401 Complete Course
Let's see an example of what happens when a parameter mismatch occurs. This short clip from Jeremy's IT Lab demonstrates the result of changing the native VLAN on just one member port.
Watch from 19:28 to 20:43. Notice how the show etherchannel summary output changes, displaying the 's' flag for the misconfigured port, effectively removing it from the active bundle.
If you see an s flag, your next step is to perform a side-by-side comparison of the running configuration for all member ports (e.g., show run interface g0/0 and show run interface g0/1) to find the discrepancy.
2.2. Protocol and One-Sided Configuration Issues
A more dangerous situation arises when the two switches on either end of the link have conflicting EtherChannel configurations.
- Protocol Mismatch: One switch is configured for LACP (
active/passive) while the other is configured for PAgP (desirable/auto). The negotiation will fail, and the ports will likely show asI(stand-alone). - Mode Mismatch: One switch is LACP
passiveand the other is alsopassive. Since neither will initiate negotiation, the channel won't form. The same is true for PAgPauto/auto. - Static (
on) Mismatch: One switch is set tomode on(static bundling), while the other is set to a negotiation protocol like LACPactive. The static side forces bundling, but the LACP side fails to negotiate, leading to an unstable state.
The most critical misconfiguration is when one switch is configured for EtherChannel (especially mode on), and the other switch has no EtherChannel configuration at all. This can create a Layer 2 loop that Spanning Tree Protocol (STP) cannot see, because STP views the bundled ports on the first switch as a single link.
EtherChannel // CCNP ENCOR 350-401 Complete Course
This clip from Jeremy's IT Lab explains exactly why a one-sided static EtherChannel configuration is so problematic and can lead to a forwarding loop.
Watch from 15:57 to 19:28. Pay close attention to the packet walk-through that demonstrates how a broadcast frame can loop back to the originating switch.
2.3. EtherChannel Misconfiguration Guard
Because of the loop risk, Cisco switches have a built-in safety mechanism called EtherChannel Misconfiguration Guard. This feature, which is part of Spanning Tree and enabled by default, detects when an EtherChannel is inconsistent.
To understand how it works, let's explore the underlying mechanism.
Understand EtherChannel Inconsistency Detection
This Cisco support document explains the theory behind the inconsistency detection. As a solutions engineer, you'll appreciate understanding the 'why' behind the feature.
Read the 'Background Information' and 'How Inconsistency Detection Works' sections. Focus on how the switch uses the source MAC address of incoming STP BPDUs to detect that it's receiving BPDUs from multiple distinct ports on the other side, which shouldn't happen on a correctly formed channel.
When the guard detects an inconsistency, it puts the affected ports on the local switch into an err-disabled state to prevent a loop.
Let's see this feature in action.
EtherChannel - Load Balancing and Misconfiguration Guard
In this demonstration from Kevin Wallace Training, LLC, an interface is deliberately removed from a channel on one switch, creating an inconsistency. You will see how the misconfiguration guard on the partner switch reacts.
Watch from 08:10 to 12:18. Observe the console messages that appear and the subsequent output of show interface status, which confirms the ports are now err-disabled.
To recover from an err-disabled state caused by this issue, you must first fix the underlying configuration mismatch on the remote switch. Then, on the local switch, you can bring the ports back up by issuing a shutdown followed by a no shutdown command on the physical interfaces.
Test your understanding!
You are troubleshooting a connection between SwitchA and SwitchB. The Port-channel is down, and show etherchannel summary on SwitchA shows the member ports Gi0/1 and Gi0/2 have the (I) flag, meaning they are stand-alone.
What are two different configuration mismatches that could cause this symptom?
Show answer
Here are two common causes for ports being in a stand-alone state:
- Protocol Mismatch: SwitchA could be configured to use LACP (
mode active), while SwitchB is configured to use PAgP (mode desirable). The protocols are incompatible and cannot negotiate a channel. - Passive/Auto Mismatch: Both switches could be configured with a non-initiating mode. For example, both are set to LACP
passive, or both are set to PAgPauto. Since neither switch will actively try to form a channel, no negotiation occurs.
3. A Structured Troubleshooting Workflow
When faced with a faulty EtherChannel, a structured approach is key.
-
Check the Port-channel Status: Start with
show etherchannel summary. Interpret the flags (P,I,s,D, etc.) to get an initial diagnosis. -
Check the Physical and Interface Status: Use
show ip interface briefandshow interfaces statusto ensure the physical member interfaces areupandconnected. An EtherChannel cannot form over down interfaces. -
Compare Configurations: Use
show run interface <interface_name>for all member ports on both switches. Meticulously compare the speed, duplex, trunking encapsulation, native/allowed VLANs, and channel-group commands. This is where most issues are found. -
Check Protocol-Specific Details: If the configurations appear to match but the channel still won't form, use protocol-specific commands to check the negotiation status. This will help you see what the switch is hearing from its neighbor.
Troubleshoot EtherChannels on Catalyst 9000 Switches
This Cisco troubleshooting guide provides a wealth of commands for deep-diving into LACP and PAgP. We'll focus on the basic verification commands.
Review the 'Basic Checks' subsections within both 'Verify LACP Operation' and 'Verify PAgP Operation'. You don't need to memorize every command, but be familiar with show lacp neighbor and show pagp neighbor. These commands are invaluable for seeing the neighbor's state from the local switch's perspective.
Using commands like show lacp neighbor can reveal the partner's system ID, port priority, and flags, which can quickly expose why negotiation is failing.
Conclusion
You have now completed the Advanced LAN Switching module by learning how to troubleshoot the technologies we've built. A stable and efficient EtherChannel relies on perfect consistency, and now you have the tools to enforce it.
Key Takeaways:
- Your troubleshooting journey almost always begins with
show etherchannel summary. Understanding its flags is non-negotiable. - The most common issues are parameter mismatches (speed, duplex, VLANs), which result in suspended (
s) ports, and protocol mismatches, which result in stand-alone (I) ports. - A one-sided configuration can create dangerous STP loops. The EtherChannel Misconfiguration Guard protects against this by placing ports in an
err-disabledstate. - A systematic workflow is essential: check the summary, check physical status, compare running configs, and then use protocol-specific commands like
show lacp/pagp neighbor.
Preview of the Next Module:
We've repeatedly mentioned how Spanning Tree Protocol (STP) interacts with EtherChannel and protects the network from loops. In our next module, we will dive deep into Spanning Tree Protocol and First-Hop Redundancy. Our first lesson will compare the modern Rapid Spanning Tree Protocol (RSTP) to its legacy predecessor, focusing on the critical topic of convergence time.