The device owner of one of the devices we monitor with Spectrum has reported that Spectrum is polling Juniper and Cisco models at a higher polling frequency than other devices of the same model and chassis. I've verified the poll interval in Spectrum and the model is configured no different than other models. I've also deleted the model and rediscovered it but the issue still persists. Spectrum is polling for ifadminstatus and ifoperstatus on many interefaces.
Component : Spectrum Core / SpectroSERVER
Spectrum will monitor the ifadmin and ifoperstatus of LACP interfaces to determine if the configured threshold is violated. This causes extra polling to the devices that support LACP:
LACP Threshold Computation
Spectrum generates various LACP threshold violation alarms on the LACP port channel model, such as critical, major, and minor. All the physical interfaces that are aggregated to the logical port of LACP participate in threshold calculation. You can use this threshold alarm to know the usage of the physical interfaces in an aggregate group of LACP.
The LACP threshold is calculated based on the IfOperStatus of the aggregated interfaces. To access IfOperStatus, you must have the admin link (based on external attribute, ifAdminStatus) status as UP.
down/totallinks
The following scenarios are considered during the threshold comparisons:
If the LACP_IF_Port attributes are set to a value other than 0 but not greater than 100, the following attributes are used during the threshold comparisons:
LACP_Critical_Threshold = 0x2202c0
LACP_Major_Threshold = 0x2202bf
LACP_Minor_Threshold = 0x2202be
If the LACP_IF_Port attributes are set to 0 or less than 0 or greater than 100, the following VNM attributes values are used during the threshold comparisons:
Critical_Threshold = 0x0001321e
Major_Threshold = 0x0001321d
Minor_Threshold = 0x0001321c
If all the LACP port level attributes are set to 0 and the VNM LACP threshold attributes are set to greater than 100, no threshold is calculated.