Custom CiscoMemoryPool attribute monitoring in Spectrum
search cancel

Custom CiscoMemoryPool attribute monitoring in Spectrum


Article ID: 233707


Updated On:


CA Spectrum DX NetOps


We have received a request to enable monitoring for the below OID's on the Cisco 4k device.  But there are no traps from the vendor to do that and these attributes do not exist in Spectrum.  How can we monitor them? 




Release : 20.2.x, 21.2.x

Component : Spectrum Alarm & Event Management


This device is not supported by Spectrum until 21.2.6 so prior to that it can be modeled as a GnSNMPdev device.  ( - Firepower 4140 SM36 )

The Cisco-Memory-Pool-Mib contains the attributes in question but they are not mapped by default.

The vendor has not provided any traps to monitor this functionality.


Firepower4k model type are only supported on 21.2.6 and are modeled as FirePower_4k_9k instead of the GnSnmpDev model type. 

To create watches for these oid on all devices of this type, you would create it on one and make it inheritable to the FirePower model type.  

*** Do not do this on the GnSnmpDev model type as all generic SNMP devices will inherit this polling which could cause performance problems for Spectrum.  A custom model type could be made on releases prior to 21.2.6***


it shows these values in the mib for this oid are in a table.

I have created and tested successfully a watcher on the ciscoMemoryPoolUsed attribute on all instances in the table as data type gauge.

in the properties tab, enable the polling as required.

configure the threshold as you wish.

The watch has been violated and creates events and alarms as follows:



Additional Information

The memorypoolused attribute value is in bytes

ciscoMemoryPoolUsed OBJECT-TYPE
 SYNTAX Gauge32 (0 .. 4294967295)
 "Indicates the number of bytes from the memory pool that are currently in use by applications on the managed device."
  ::= { ciscoMemoryPoolEntry 5 } --

To create a threshold on percentage of memory used, you would need an equation to add used and free memory of each instance, to get total memory and then calculate the used memory as a percentage of that total.