High Buffer utilization on FPC (Flexible PIC Concentrator)
The problem related to this syslog message is described in the following sections:
The RMON_EVENT_fpc_high_buffer_utilization message contains information about an FPC that has a high buffer utilization rate.
An alarm is triggered when the FPC buffer utilization crosses a threshold.
High FPC buffer utilization can be caused by high traffic volume between the Packet Forwarding Engine and the Routing Engine, and usually doesn’t represent any kind of problem; but it can also be caused by a memory leak condition.
The output from show chassis fpc will show current utilization levels. This is useful to show the levels over time, so you can determine if the high utilization levels are transitory. Different types of FPCs will allocate buffer space in different ways. This means that you can expect the buffer utilization level to vary from FPC to FPC. Additionally, this value only tells you how much of the buffer space is allocated for use – not how much is actually being used. This means that you may see buffer utilization levels that seem high, even if the FPC is not moving any traffic at all.
Allocated buffer space that is not actually being used will be re-allocated as needed by internal processes that use the FPC’s buffer.
Monitor the FPC using show chassis fpc. It is not unusual to see periods of high buffer utilization. Check the messages log for ‘packet allocation failure’ messages. If there are none present and there are no other symptoms other than high buffer utilization, then no action is necessary.
A memory leak condition will usually affect the FPC’s memory before it affects the buffer allocation, and the memory loss will be sustained rather than transitory.
Higher than expected buffer usage can sometimes be resolved by restarting the FPC. If this does not help, or if you see memory utilization problems, packet allocation failure messages in the syslog, or other symptoms that you think might be caused by the high buffer utilization, collect the following outputs:
show log messages show chassis fpc
request pfe execute command “show syslog messages” target fpc0 (for this command, replace 0 with the number of the fpc that is experiencing high buffer utilization)
request pfe execute command “show nvram” target fpc0 (for this command, replace 0 with the number of the fpc that is experiencing high buffer utilization)