7.2.8 Planning the data to be collected
The State Poller service uses polls to gather status information about the monitored devices in your network. Polling can be done using ICMP, SNMP, or both.
- ICMP (ping)
ICMP address monitoring uses ping requests to verify the availability of each managed IP address.
- SNMP Polling
SNMP monitoring verifies that each monitored SNMP agent is responding to SNMP queries.
The State Poller is highly optimized to collect configured SNMP information from each monitored object with one query at each interval. When you save configuration changes, the State Poller recalculates the group membership of each object and reapplies the configured interval and set of data to collect.
SNMP monitoring issues SNMP queries for all monitored interfaces and components, requesting the current values from the MIB II interface table, the HostResources MIB, and vendor-specific MIBs. Some values are used for fault monitoring.
- Web Polling
-
If Web Agents are configured (in addition to SNMP Agents), NNMi can use additional protocols. For example, SOAP protocol for VMware environments.
- SNMP component health data
You might enable or disable component health monitoring at the global level. Component health monitoring for faults follows the fault polling interval settings for the device.
Gathering additional data at each poll does not affect the time to execute the poll. However, additional data stored for each object can increase the memory requirements for State Poller.
- Tip
Batching your monitoring configuration changes is less disruptive to State Poller ongoing operation.