Troubleshoot StackWise on Catalyst 9200/9300: Fix Half Ring, Stack Port Down, and Reload Loops
StackWise Not Working? Quick Answer
If your Cisco Catalyst 9200 or 9300 StackWise ring is down or unstable, the fastest way to isolate the fault is to run show switch stack-ring speed and show switch stack-ports summary. A healthy stack must operate in a Full ring state with all ports showing LinkOK. If you see a Half ring or ports marked as Down, immediately check for a loose, damaged, or incorrectly crossed stack cable. Reseating or replacing the faulty StackWise cable resolves the majority of split stack and reload loop issues.
The Frustration of a Broken Stack
A switch stack is supposed to provide seamless redundancy and simplified management. But when a StackWise cable degrades or an internal stack-manager process crashes, the exact opposite happens. You might face constant active/standby re-elections, a “split-brain” scenario where multiple switches try to act as the master, or sudden reload storms that take down your entire access layer.
You do not need to reboot switches blindly to fix this. In this guide, we provide a definitive engineering runbook to troubleshoot StackWise failures on Cisco Catalyst 9200 and 9300 series switches. We will cover the exact CLI commands to use, how to interpret specific syslog errors, and provide a clear decision tree to help you isolate physical Layer 1 faults from IOS-XE software bugs.
Grab your console cable, verify your stack ring state, and follow the diagnostic playbooks below to restore your backplane to a stable, Full ring topology.
Common Cisco StackWise Failure Symptoms (Half Ring, Ports Down)
When a Catalyst stack fails, the symptoms point directly to the underlying hardware or software fault. Here is a quick diagnostic matrix:
- Half Ring: One stack cable is completely disconnected, damaged, or unseated. The stack is functioning, but physical redundancy is broken, and backplane bandwidth is halved.
- Stack Ports Down: The physical link between two switches has failed. This points to a dead cable, bent pins on the connector, or a stack module that isn’t screwed in flush with the chassis.
- Member Not Joining (State: V-Mismatch / Provisioned): The new switch has a different Cisco IOS-XE version, a different license tier (e.g., Network Essentials vs. Advantage), or is booting in Bundle mode while the stack is in Install mode.
- Split Stack / Master Role Changes: Intermittent stack cable flaps are causing the switches to lose sight of each other, triggering constant active/standby re-elections.
- Stack Keeps Reloading: Often caused by a severely degraded stack cable injecting cyclic redundancy check (CRC) errors into the ring, leading to stack-manager crashes.
(Note: The ultimate goal of your troubleshooting is to restore the stack to a Full ring state).
Catalyst 9200 vs 9300 StackWise Bandwidth Limits
Understanding the hardware limits of your specific platform is critical, especially when troubleshooting performance bottlenecks during a Half Ring failure.
Cisco Catalyst 9200 and 9300 switches use different generations of StackWise technology. The cables and stacking modules are not interchangeable between the two series.
| Platform | StackWise Architecture | Maximum Stacking Bandwidth | Impact of a “Half Ring” Failure |
| Catalyst 9200 | StackWise-160 | 160 Gbps | Bandwidth drops to 80 Gbps |
| Catalyst 9300 | StackWise-480 | 480 Gbps | Bandwidth drops to 240 Gbps |
| Catalyst 9300X | StackWise-1T | 1 Tbps (1000 Gbps) | Bandwidth drops to 500 Gbps |
The Engineering Impact: When show switch stack-ring speed reports a Half ring, you do not just lose hardware redundancy; your backplane capacity is instantly cut by 50%. In high-throughput environments (like heavy video surveillance or Wi-Fi 6/6E aggregation), this sudden bandwidth drop can cause dropped packets and severe network congestion.
StackWise Troubleshooting Decision Tree
To stop guessing and start fixing, follow this logical progression to isolate the fault. Begin by determining the ring integrity, then check port health, and finally look for system instability.
代码段
flowchart TD
A[StackWise issue reported] --> B{show switch stack-ring speed}
B -->|Full| C{show switch stack-ports summary}
B -->|Half| H[Inspect/ reseat/ swap stack cables\nverify neighbor mapping]
C -->|LinkOK stable| D{Any syslogs of reloads or role changes?}
C -->|LinkOK flaps| I[Suspect cable/connector/port\nreseat + swap known-good]
D -->|No| E[Validate config/compatibility\nthen monitor]
D -->|Yes| F[Follow reload due to stack issues workflow\ncollect crashinfo + isolate member]
F --> G[Stabilize member: reseat hardware\nverify IOS-XE code alignment]
Essential CLI Commands to Verify Cisco Stack Status
Instead of reloading the stack and hoping the problem disappears, use these five commands to pinpoint exactly where the ring is breaking down.
1. show switch
Provides a high-level overview of the stack members, their roles (Active, Standby, Member), and their current state.
- Healthy Signal: All connected switches should show a state of Ready. States like Provisioned or V-Mismatch indicate the switch is physically connected but logically isolated due to a software mismatch.
2. show switch detail
Expands on the basic command by showing the hardware type, MAC addresses, and stack port status for each specific member.
3. show switch neighbors
Maps out the exact physical topology of the stack. It shows which switch is connected to Port 1 and Port 2 of any given member.
- Healthy Signal: The logical output must perfectly match your physical cabling. Switch 1 should connect to Switch 2, Switch 2 to Switch 3, and the final switch must loop back to Switch 1.
4. show switch stack-ports summary
Checks the low-level physical health of the StackWise cables and interfaces.
- Healthy Signal: You want to see LinkOK and SyncOK for every active port. If you see Down or NoCable, the physical link is broken.
5. show switch stack-ring speed
Confirms the overall bandwidth and redundancy of the backplane.
- Healthy Signal: The output must state Configuration: Full. If it says Half, your stack has a broken path.
How to Fix Catalyst Stacking Issues (Half Ring & Reloads)
Half Ring After a Cabling Change
If your ring drops to a Half state after adding a switch or replacing a cable, the physical loop is broken. First, verify the cabling logic: Port 1 on a switch should ideally connect to Port 2 on the adjacent switch. If the cables are crossed correctly, physically inspect the connectors. A StackWise cable that isn’t screwed in completely straight will fail to link. Reseat the cables on both ends of the “Down” connection identified by show switch stack-ports summary.
Member Won’t Join / Version Mismatch
When show switch reveals a state of V-Mismatch, the hardware is fine, but the software is blocking the join process. Ensure the new switch is running the exact same IOS-XE version and license level as the active switch. Furthermore, Catalyst 9200/9300 stacks strongly prefer Install mode. If the new switch is booting in Bundle mode (.bin file), it will fail to auto-upgrade and join the stack.
Intermittent Stack Port Flaps (LinkOK Changes)
If a stack port is bouncing between LinkOK and Down, your stack will experience massive performance hits and potential split-brain scenarios. This is almost always a Layer 1 hardware issue.
Syslog Verification: Check
show logging. If you see the error message%STACKMGR-4-STACK_LINK_CHANGE: Stack Port 1 Switch 2 has changed to state DOWN, this is your definitive proof of a physical layer fault.
Completely power down the affected member, remove the stack module (if using a 9200/9300 modular chassis), ensure it is seated flush, and replace the stack cable with a known-good spare.
Unexpected Reloads Attributed to Stack Issues
A Catalyst stack that keeps reloading is often the victim of an unstable StackWise link injecting continuous errors into the backplane, causing the stack-manager process to crash. Look at the show logging or show crashinfo output just before the crash for interface flapping on the stack ports. If a specific member is constantly crashing, physically isolate it from the ring to see if the rest of the stack stabilizes.
Cisco Stack Troubleshooting Runbook Table
Use this quick-reference table during active troubleshooting to validate your stack’s health in under 60 seconds.
| Step | Goal | Command | Healthy Signal |
| Verify ring | Full vs Half | show switch stack-ring speed | Configuration: Full |
| Find bad link | Cable/port health | show switch stack-ports summary | LinkOK / SyncOK stable |
| Confirm topology | Neighbor mapping | show switch neighbors | Matches physical cabling |
| Check instability | Reload reasons | show logging / dir crashinfo: | No stack-triggered reload loops |
People Also Ask (FAQ)
What does Half ring mean on Catalyst StackWise?
A Half ring means that one of the stack cables is disconnected, damaged, or failing. The switches can still communicate, but the stack has lost its redundant loop, and the total backplane bandwidth is cut in half (e.g., from 480 Gbps down to 240 Gbps on a 9300).
Which command shows StackWise ring status?
The command show switch stack-ring speed displays whether the stack is operating in a Full or Half ring state. You can also use show switch stack-ports summary to see the physical status of individual stack ports.
Why does a Catalyst stack keep reloading?
Reload loops in a stack are typically caused by a severely degraded stack cable causing constant active/standby re-elections, a loose backplane stack module, or an IOS-XE software bug crashing the internal stack-manager process.
Can I hot-swap a StackWise cable if the ring is Full?
Yes. If show switch stack-ring speed confirms the stack is in a Full ring state, you can safely remove and replace one stack cable without bringing down the switch, as the traffic will instantly route over the remaining half of the ring.
What’s the fastest way to find the unstable member?
Run show switch stack-ports summary and look for any ports that are flapping or listed as Down. The switch with the problematic port is the source of the instability, pointing you directly to the bad cable or module that needs to be reseated.
