I monitor the system health of a Cisco ASA via SNMP. Unfortunately, there occurs a CPU hog on the ASA which I cannot explain. Is it possible that ASA monitoring with PRTG somehow causes this issue with my ASA?


Article Comments

This article applies to PRTG Network Monitor 15 or later

CPU Hogs on Cisco ASA and SNMP Monitoring

If you monitor memory on a Cisco ASA with the SNMP Cisco System Health sensor, you might encounter excessive CPU load on the Cisco ASA device. The reason for this might be a known issue with the Cisco ASA and the OID the sensor uses to retrieve memory usage: The SNMP polling of the ASA memory pool information causes CPU hogs.

The SNMP Cisco System Health sensor queries the MEMPOOL_GLOBAL_SHARED pool via SNMP to monitor memory usage (the corresponding OID is 1.3.6.1.4.1.9.9.48.1.1.1.X.7). This SNMP process might hold the CPU of the ASA for too long before freeing it for other processes which results in a CPU hog. Another result of this issue might be that packets are dropped if the data rate is too high through the ASA.

Unfortunately, we cannot control this behavior because this is a bug on the Cisco ASA device. Please see the according Cisco bug report for details (login required).

We recommend that you pause the SNMP Cisco System Health sensor that monitors memory usage on your Cisco ASA for as long as the bug exists to avoid CPU hogs and packet loss.


Jul, 2015 - Permalink