Some Enterasys devices cannot handle the amount of SNMP requests received from the SpectroSERVER. This is causing the cpu to spike on these devices and possibly shutdown. The devices are able to be initially discovered in the Spectrum database. It is when the discovery process reads large tables like the bridge table to determine connectivity which causes the cpu to spike.
Enabling Throttling on the device or the model type may help.
Throttling - Attribute ID: 0x11f79 - Controls whether CA Spectrum should restrict the amount of outstanding SNMP requests sent to a device. This helps to alleviate problems involving SNMP agents that cannot handle large amounts of SNMP requests.
Throttle_Count - Attribute ID: 0x11f39 - Determines how many outstanding SNMP requests are allowed when throttling is enabled for a device.
The Throttling and Throttle_Count attributes can be modified in the Attributes tab of the individual Enterasys model. The Attribute Editor can also be used to modify these attribute for the individual model and to set the default values for the model type.
If enabling Throttling does not help, disable automatic discoveries on the Enterasys model by setting the "Discover Connections After Link Up Events" and "Discover After Reconfigure" to "No" in the Reconfiguration folder in the Information tab to help minimize the amount of snmp sent to the device for discovery purposes. However, doing any one of the following may still cause large amounts of SNMP to these Enterasys devices.
- Running an AutoDiscovery Discovery and Modeling configuration
- Right mouse clicking on the Enterasys model and selecting Reconfiguration -> Discover Connections
- Right mouse clicking on a container model that contains an Enterasys model and selecting Reconfiguration -> Discover Connections
- Running an automatic or manual rediscovery of a device model that is connected to an Enterasys model
Discovery and model configuration can be configured to not read bridge tables. However, the connectivity will need to be done manually.
The vendor should be contacted to report this issue.