Field Notice: FN74354 - Cisco MDS 9000 Series Fabric Switches Running Cisco NX-OS Software Release 9.4(4) May Experience Unexpected System Reload - Software Upgrade Recommended

Cisco MDS 9000 Series Fabric Switches that are running Cisco NX-OS Software Release 9.4(4) may experience an unexpected system reload when the management interface (mgmt0) experiences an ungraceful link failure or flapping condition. This issue affects specific models in the Cisco MDS 9000 Series product line and can result in service disruption.

A software defect, CSCws09730, was introduced in Cisco NX-OS Software Release 9.4(4) that causes the port service to crash when the management interface (mgmt0) link state ungracefully changes to down. This defect can lead to repeated service crashes and ultimately trigger a disruptive automatic system reload based on the High Availability (HA) policy.

When this issue occurs, one or more of the following symptoms may be observed:

Primary Symptoms

1. Unexpected Switch Reload - The switch experiences an unplanned reload, causing service disruption.

If the condition leads to a reload, the system logs will show repeated crashes of the port service:

%SYSMGR-2-SERVICE_CRASHED: Service "port" (PID 3920) hasn't caught signal 11 (core will be saved).
%SYSMGR-2-SERVICE_CRASHED: Service "port" (PID 32356) hasn't caught signal 11 (core will be saved).
%SYSMGR-2-SERVICE_CRASHED: Service "port" (PID 32403) hasn't caught signal 11 (core will be saved).
%SYSMGR-2-SERVICE_CRASHED: Service "port" (PID 32472) hasn't caught signal 11 (core will be saved).

After the reload, the show system reset-reason command will indicate that the reset was triggered by the HA policy for the port service:

----- reset reason for module 1 (from Supervisor in slot 1) ---
1) At 217995 usecs after Wed Nov 12 18:22:48 2025
    Reason: Reset triggered due to HA policy of Reset
    Service: port hap reset
    Version: 9.4(4)

2. Garbled Syslog Messages - Console or syslog output displays corrupted characters when the mgmt0 interface goes down:

   %PORT-5-IF_DOWN: Interface mgmt0 is down () \201^U^[

Trigger Condition

This issue is triggered when the management interface (mgmt0) link goes down ungracefully or flaps, typically due to one or more of the following:

• Management network connectivity issues
• Peer management switch failures
• mgmt0 cable disconnection or failure
• Planned or unplanned management network maintenance

Bringing the link down gracefully using a configuration shutdown of the mgmt0 interface from the MDS side does not trigger the issue.

Cisco has identified the following workarounds and solutions for this issue:

Recommended Solution

Upgrade to a fixed software release: This issue will be fixed in the upcoming Cisco NX-OS Software Release 9.4(5) (available in Q1 2026) and later releases.

Workaround

If immediate upgrade is not possible, downgrade to Cisco NX-OS Software Release 9.4(3b), which does not contain this defect.

Additional Mitigation

To minimize the risk of triggering this condition while running affected NX-OS code, do the following:

• Ensure management network infrastructure is stable and redundant where possible.
• Monitor management interface connectivity.
• Plan management network maintenance carefully to avoid unnecessary link flaps.

Important Notes

• Cisco NX-OS Software Release 9.4(4) has been deferred for the affected products due to this defect.
• Cisco MDS 9700 Series Director-class Switches (9706, 9710, and 9718) are not affected by this issue.

To identify the product ID (PID) of a device, use the following command:

switch# show inventory | head lines 2
NAME: "Chassis",  DESCR: "MDS 9148V 48X64G FC (1 RU) Chassis"
PID: DS-C9148V-K9        ,  VID: V01 ,  SN: JPG2631008X

To check which Cisco NX-OS Software release is running on a device, use the show version command:

switch# show version
...
system:    version 9.4(4)
...

If Cisco NX-OS Software Release 9.4(4) is running on any of the affected product models listed above, the switch is considered vulnerable.