Re: [squid-users] Client bypassing delay pool restrictions

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Tue, 16 Nov 2010 02:40:33 +1300

On 15/11/10 20:05, RM wrote:
> Hello all,
>
> I am running Squid Cache: Version 2.6.STABLE21 on CentOS 5.5 and have
> been using delay pools to limit clients' bandwidth usage. Here is the
> delay pool section and related ACL of the squid.conf file. I have
> included the entire squid.conf at the end of the message:
>
> acl all src 0.0.0.0/0.0.0.0
> delay_pools 1
> delay_class 1 1
> #1Mbps
> delay_parameters 1 131072/131072
> delay_access 1 allow all
>
> I have used the above delay pool configuration countless times
> previously and I did not have any issue but for some reason there is a
> client that is able to bypass the delay pool bandwidth restriction and
> transfter at rates of 5Mbps+.
>
> Any help would greatly be appreciated.
>
> Thanks in advance!
>
> Ron M.
>

More likely that those requests are ones where the client actualy
disconnected. Your quick_abort setting configure Squid to keep going
after a disconnect. This happens outside the pooling since there is no
client to pool.

It *might* be a client doing some tricky request behaviour. You could
pick this up by a) these requests are *all* MISS requests (indicating
only-of-cached header preventing slow network access), or b) these
requests following an earlier request within a very short time
(indicating a leach re-attachment once the above pool detachment has
been done by Squid).

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE9 or 3.1.9
   Beta testers wanted for 3.2.0.3
Received on Mon Nov 15 2010 - 13:40:37 MST

This archive was generated by hypermail 2.2.0 : Tue Nov 16 2010 - 12:00:03 MST