Upgrade Guide for Cisco Catalyst 9000 Series Switches and Smart Switches

PDF

Upgrade Guide for Cisco Catalyst 9000 Series Switches and Smart Switches

Recommendations for device health check

Updated: August 1, 2026

Want to summarize with AI?

Log in

Outlines recommendations for determining device health, including pass criteria and indicators for further analysis based on install state, redundancy, stack and module status, resource utilization, counters, network recovery, logs, applications, and service confirmation.

Pass criteria

A device passes the health check when the intended release is running and properly activated, all redundancy and stack members are healthy, resource use remains within normal limits, counters are stable, expected interfaces and applications have recovered, logs contain no critical errors, and normal service is confirmed.

The pass criteria includes these:

  • Pass criteria - install and redundancy state

    • The intended release is running and packages.conf file is used in install mode.

    • The image is activated and committed, with no active rollback timer.

    • No ISSU transaction remains in progress.

    • SSO is operating, communications are up, and the peer is STANDBY HOT where applicable.

    • Configuration-synchronization failure lists are empty.

  • Pass criteria - platform and stack health

    • Every expected stack member, supervisor, line card, SVL, and DAD link is healthy.

    • CPU, memory, load, fans, power, and temperature remain healthy.

    • Error, discard, stack-link, and redundancy counters are stable between samples.

  • Pass criteria - network, logs, and application recovery

    • Expected interfaces, port channels, routes, neighbors, FHRP groups, and hosted applications have recovered.

    • Logs contain no unexplained crash, traceback, critical platform, install, or ISSU messages.

    • Application owners and monitoring systems confirm normal service.

When to investigate further

Initiate further investigation if the device reports warning or critical statistics, experiences persistent high CPU or declining memory, shows discrepancies in image or peer status, has incomplete install or ISSU transactions, reveals missing or unhealthy stack and links, registers increasing errors, fails to restore network baselines, or presents unexpected log incidents.

  • Install, redundancy, and stack indicators

    • The active and standby processors run different images or the peer is not STANDBY HOT.

    • An ISSU or install transaction remains pending, incomplete, failed, or rolled back.

    • A stack member is missing, not Ready, or the physical ring reports Half.

    • SVL or DAD links are down, suspended, pending, or not ready.

  • Interface, network, and application indicators

    • Interface errors, discards, stack CRC errors, or redundancy errors continue increasing.

    • Routing neighbors, route counts, FHRP groups, or hosted applications do not return to their baseline state.

    • An unexpected reload, crash, core file, traceback, or critical log message is present.