Cisco ASR 9000 show environment: Power, Fans and Temperature

Hardware problems on an ASR 9000 rarely announce themselves politely. A failing power supply, a fan tray spinning down or a slot running hot shows up as an intermittent line card reset or a sudden "not enough power" message during a routine reload. The show environment family, plus show platform and show memory summary, is how you see those conditions before they become outages.

show environment: the physical picture

RP/0/RSP0/CPU0:router# show environment
RP/0/RSP0/CPU0:router# show environment ?

The command reports fan speeds and states, power supply voltage and current, and temperature readings per component. Read it in two passes: first the states (any fan or power supply reporting a state other than OK), then the temperatures against their thresholds. A fan tray that has dropped from its normal RPM while still reporting OK is a common early warning — compare the numbers across fan trays rather than accepting the per-tray verdict.

Power budgeting

Power matters most when the chassis is under-populated or when a module refuses to come up. Before blaming the module, confirm that the available power budget covers the installed cards and any PoE or line-card startup draw, and check the power-supply redundancy state. A supply that reports present but is not contributing is a chassis running without redundancy — a finding worth a change record on its own.

RP/0/RSP0/CPU0:router# show environment power
RP/0/RSP0/CPU0:router# show platform
RP/0/RSP0/CPU0:router(admin)# show platform

show platform in administration EXEC toggles between the whole router and the nodes of the current Secure Domain Router, and every node should report IOS XR RUN. A card stuck in a state other than RUN is either still initialising, failed to program, or is being held back by power or FPD mismatches.

Memory and the low-memory failure mode

RP/0/RSP0/CPU0:router# show memory summary
RP/0/RSP0/CPU0:router# show watchdog
RP/0/RSP0/CPU0:router(config)# watchdog threshold memory location 0/RP0/CPU0 minor 30 severe 20 critical 10

IOS XR supports configurable watchdog thresholds per node. When memory reaches the severe or critical state, normal users cannot enter configuration mode and only lr-root owners can make changes — and even they may need commit force to apply a configuration that reduces memory usage. Raising the minor/severe/critical thresholds from their defaults (20/10/5) buys earlier warning on an RP that is trending upwards, but it does not fix a leak: collect show processes memory and show watchdog memory-state before assuming you know the cause.

A five-minute health sweep

RP/0/RSP0/CPU0:router# show environment
RP/0/RSP0/CPU0:router# show platform
RP/0/RSP0/CPU0:router# show redundancy
RP/0/RSP0/CPU0:router# show memory summary
RP/0/RSP0/CPU0:router# show install committed
RP/0/RSP0/CPU0:router# show logging | include %PLATFORM

This sequence answers the questions every TAC case begins with: is the hardware healthy, are all nodes running, is the standby RP in sync, is memory trending, what software is committed, and has the platform logged a fault that the CLI summary does not surface.

Related reading: ASR 9000 IOS XR show commands, ASR 9000 fabric troubleshooting, IOS XR configuration management.

原文链接：Troubleshooting the Cisco IOS XR Software