Re: [squid-users] squid cache manager question and snmp with smp question

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Tue, 12 Nov 2013 22:11:07 +1300

On 12/11/2013 9:21 p.m., Dr.x wrote:
> hi ,
> from cache manager :
> Cache information for squid:
> Hits as % of all requests: 5min: 11.7%, 60min: 11.0%
> Hits as % of bytes sent: 5min: 0.6%, 60min: -0.3%
> Memory hits as % of hit requests: 5min: 20.0%, 60min: 13.9%
> Disk hits as % of hit requests: 5min: 11.6%, 60min: 9.5%
> Storage Swap size: 28703904 KB
> Storage Swap capacity: 70.1% used, 29.9% free
> Storage Mem size: 1024000 KB
> *Storage Mem capacity: 100.0% used, 0.0% free*
> Mean Object Size: 32.00 KB
> Requests given to unlinkd: 0
> ================================================
> im asking about :
> Storage Mem capacity: 100.0% used, 0.0% free
> why it is 100 %
>

Because your cache is busy and using all the memory you gave it for
caching objects (cache_mem). This is normal for memory cache, there is
no reason to hold it lower than 100% since there is no delay in deleting
things when they need to be.

> Q1-does that mean that squid dissipated 100% from cache_mem value configured
> ????

"dissipated"? no, used.

> Q2- does the result in cache manager in "general run time information" is
> calculated as total for all processes ???

Um. Is cache_mem set to 1000 MB in on worker? or would that be the sum
for all the workers with cache_mem ?

> ==================================================================
> Q3
> ABOUT snmp with smp
> wt i need to configure in squid.conf ??

Squid must be built with --enable-snmp.

Also, snmp_port and snmp_access directives must be configured.

>
> do i need to configure snmp for each instance ???

http://wiki.squid-cache.org/Features/SmpScale#What_can_workers_share.3F

>
> i want to say that i configured as below :
> acl snmppublic snmp_community xxxxxxx
> snmp_port 3401
> snmp_access allow snmppublic localhost
> snmp_access allow snmppublic all
> snmp_incoming_address 0.0.0.0
> snmp_outgoing_address 0.0.0.0

Drop those last two lines about address. The first one is doing nothing
useful. The second one will cause failures.

From the config manual:
"
  NOTE, snmp_incoming_address and snmp_outgoing_address can not have
        the same value since they both use the same port.
"

> ###############################################
>
> i had results in my mrtg , but not sure of the results ,
> i got squid mib file and converted it to oidb file and put it in m y mrtg
> .
>
> by here im using smp , not sure from the results
>
>
> i also revived some suspicious logs :
>
> 2013/11/08 16:51:26 kid3| snmpHandleUdp: FD 20 recvfrom: (11) Resource
> temporarily unavailable

We are still trying to figure this one out. It seems not to be harmful
particularly, except a waste of effort somewhere.

Amos
Received on Tue Nov 12 2013 - 09:11:17 MST

This archive was generated by hypermail 2.2.0 : Tue Nov 12 2013 - 12:00:06 MST