Re: [squid-users] Monitoring multiple squid instances with snmp

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Thu, 13 Jan 2011 01:17:10 +1300

On 13/01/11 00:29, Artemis BRAJA wrote:
>
> On 01/12/2011 12:00 PM, Amos Jeffries wrote:
>> On 12/01/11 04:37, Artemis BRAJA wrote:
>>> Thanks Amos!
>>> But now I'm facing another problem while trying to monitor the
>>> percentage use of the CPU with the 1.3.6.1.4.1.3495.1.3.1.5.0 OID.
>>> While polling (v1 or v2), it keeps returning only the value "1" (each
>>> instance).
>>> As I saw on the wiki http://wiki.squid-cache.org/Features/Snmp, this OID
>>> is for squid version 2.2+, and I'm currently experimenting with version
>>> 3.2.0.4.
>>> Is this related to the version, or am I doing something wrong?
>>>
>>> snmpwalk -c community -v 1 localhost:3401 1.3.6.1.4.1.3495.1.3.1.5.0
>>> SNMPv2-SMI::enterprises.3495.1.3.1.5.0 = INTEGER: 1
>>>
>>> Artemis
>>>
>>
>> Could be a few things.
>>
>> * the docs may be broken (unlikely). Try without the trailing .0 to
>> see if that looks right.
> I tried without .0 but no diference.
>> It's a percentage value being cast to an INTEGER for SNMP. So...
>> * it may be that you are only using 1% or 100% of a CPU by Squid.
>> Check this against other software displaying the machines CPU.
> I was doing some benchmarking, so I tried with "top" and the value was
> around 15-20 % per instance.
>> * the casting may be broken and rounding a large decimal up.
>>
>> Amos
>
> Anyway I'll solve this with some shell scripts.

Thanks for the feedback. Looks like a bug. Would you mind reporting this
in bugzilla so we don't miss it when the SNMP agent gets its next updates?

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE9 or 3.1.10
   Beta testers wanted for 3.2.0.4
Received on Wed Jan 12 2011 - 12:17:17 MST

This archive was generated by hypermail 2.2.0 : Wed Jan 12 2011 - 12:00:02 MST