Evaluating Unix/Linux Load Average
The load average is the average system load over a period of time. It is conventionally given as three numbers that represent the system load during the last one, five, and fifteen minute periods.
Monitoring: Use the following counters to monitor load average:
|
SNMP OID: .1.3.6.1.4.1.2021.10.1.3.1
.iso.org.dod.internet.private.enterprises.ucdavis.laTable.laEntry.laLoad
Description: The 1 minute load average.
|
|
SNMP OID: .1.3.6.1.4.1.2021.10.1.3.2
.iso.org.dod.internet.private.enterprises.ucdavis.laTable.laEntry.laLoad
Description: The 5 minute load average.
|
|
SNMP OID: .1.3.6.1.4.1.2021.10.1.3.3
.iso.org.dod.internet.private.enterprises.ucdavis.laTable.laEntry.laLoad
Description: The 15 minute load average.
|
How to monitor Unix/Linux Load Average using MoniTiL
Prerequisite: Ensure the device to be monitored is in "Monitored Devices"
- Go to "Device Metrics" > "New"
- Template Type: "Snmp"
- Template Name: "UCD.laLoad.5min"
- Device: select the device to be monitored
- Frequency: The default frequency (5 min) is recommended
- "Save"
- The Load Average is now monitored and a graph will automatically be generated. The graph can be viewed from a daily, weekly, monthly or annual basis to assist in observing trends.
- Alerts can also be set should "UCD.laLoad.5min" fail or threshold exceed a set limit, example if Load Average > 2 for 12 concurrent 5 minute periods (1 hour).
Reference