Provide snmp calculated item for total CPU utilization

[brdr]

Hi Zabbix,

I would like to recommend that Zabbix add a calculated item in template 'Template SNMP Processors' for total CPU utilization (not load) across a server.

Presently (v2.2.3) Zabbix provides a Discovery Rule called 'Processors' in this template with an Item Prototype named 'Utilization of processor #$1'. Using this template our 16 CPU linux server is capturing/displaying data for each CPU. That is all well and fine, but a much more useful item is the total average CPU utilization across all the CPUs. I would think that this item would have been included in Zabbix.

I have a thread (45408) out there asking for help to create a calculated item for this total cpu utilization across the server for SNMP.

If I'm missing something then forgive me.

Thanks,
Don

[brdr]

If anyone knows of an OID for Linux for total cpu utilization across the server then please share.

[lcbowen3]

SNMP Total CPU utilization

Depending on the SNMP running on the machine, you can generally use these:

UCD-SNMP-MIB::ssCpuUser.0 = INTEGER: 0
UCD-SNMP-MIB::ssCpuSystem.0 = INTEGER: 0
UCD-SNMP-MIB::ssCpuIdle.0 = INTEGER: 99

If you want total usage, you can use a calculated type to add up the values.
There are also raw values available:
UCD-SNMP-MIB::ssCpuRawUser.0 = Counter32: 2278
UCD-SNMP-MIB::ssCpuRawNice.0 = Counter32: 1395
UCD-SNMP-MIB::ssCpuRawSystem.0 = Counter32: 6826
UCD-SNMP-MIB::ssCpuRawIdle.0 = Counter32: 3383736
UCD-SNMP-MIB::ssCpuRawWait.0 = Counter32: 7629
UCD-SNMP-MIB::ssCpuRawKernel.0 = Counter32: 0

[ptera]

An example

I am fighting with this issue myself - I cloned the Template SNMP processors and then changed the OIDs to the one listed here. I get nothing or errors.

If I create a graph as part of the discovery process it always says there already is a graph by that name.

If I try to create the graph on the host there are no items to add to the graph.

If I try to create the items manually it says the item is not supported.

I ran smtpwalk from the zabbix server to the host and get values returned so I know I have no firewall issues and that snmpd is running on the host.

frustrated

[kloczek]

If anyone knows of an OID for Linux for total cpu utilization across the server then please share.

Why are you using SNMP agent to monitor Linux if zabbix agent is more lighter and way effective and simpler?

[ptera]

Why are you using SNMP agent to monitor Linux if zabbix agent is more lighter and way effective and simpler?

The linux box is not a zabbix agent and the company who designed our CMS does not want other software installed on the server.

[kloczek]

Why are you using SNMP agent to monitor Linux if zabbix agent is more lighter and way effective and simpler?

The linux box is not a zabbix agent and the company who designed our CMS does not want other software installed on the server.

So "Use support Luke". If you have some data incorrectly exposed over some MIB you need to raise SR to fix it.

BTW: Linux as a typical Unix family member provides good enough proceeses separation so such claim is kind of b*locks.

[mySIGNATURE]

If anyone knows of an OID for Linux for total cpu utilization across the server then please share.

try this input this "view systemonly include 1.3.6.1.2.1.2" in Access Control