Performance statistics
Overview
Performance statistics provide categorized information about network throughput, and are obtained using SNMP. The information in a performance statistics record includes the following:
-
Suspect indicator that indicates a problem with the collected statistics; for example, when an NE is unresponsive to SNMP requests
When the Suspect indicator in a record is set, the counters in the record contain the following values:
After a suspect record collection, the next non-suspect record may contain a large value that is seen as a dramatic increase. Such an increase skews the data value for the interval and may result in misinterpretation. To alleviate this, an operator can configure a System Preferences parameter to increase, rather than reset, the Periodic Time value after the collection of a suspect record. For example, for a 15-minute collection interval, if one record is suspect and the next is not, the periodic time of the second record is recorded as 30 minutes. In this way, the NFM-P presents the first non-suspect record collected after a collection failure using a more realistic average value. See the NSP System Administrator Guide for information about setting system preferences.
Table 10-2: Performance statistics record properties
Table 10-3: Performance statistics data types and examples
NE data type |
NFM-P data type |
Example |
---|---|---|
Counter32 |
long |
32-bit counter such as Received Octets in the Interface Stats group |
Gauge32 | ||
Unsigned32 | ||
TimeStamp | ||
Counter64 |
UINT128 |
64-bit counter such as Received Broadcast Packets in the Interface Additional Stats group |
Integer32 |
double |
Radio output power, in tenths of dBm; NFM-P converts to dBm |
Integer32 |
float |
Port output power, in tenths of µW |
Integer32 |
int |
Small value, such as index, enumerator, or identifier |
TruthValue |
Boolean |
State indicator |
various |
String |
MAC address, object name |