On 11/10/2013 6:57 a.m., Nicolas Pagliaro wrote:
Hi, I am trying to get this work but I cant.
Here is my configuration and tests:
I compiled squid like this:
./configure --enable-auth-basic="basic"
There is no helper callled "basic" bundled with Squid...
--enable-auth-ntlm"ntlm"
... and there is no helper called "ntlm" bundled with Squid...
... are you embedding your own custom helpers into the Squid soruce tree?
--enable-external-acl-helpers="wbinfo_group" --enable-delay-pools
Then I use this squid conf:
cache_effective_user squid
cache_effective_group squid
cache_dir ufs /var/spool/squid/ 900 16 256
http_port 3128
coredump_dir /usr/local/squid/var/cache/squid
refresh_pattern ^ftp: 1440 20% 10080
refresh_pattern ^gopher: 1440 0% 1440
refresh_pattern -i (/cgi-bin/|\?) 0 0% 0
refresh_pattern . 0 20% 4320
acl lan src 192.168.0.0/24
http_access allow lan
visible_hostname SQUID-PROXY
delay_pools 1
delay_class 1 3
delay_access 1 allow lan
delay_access 1 deny all
delay_parameters 1 1250/1250 -1/-1 1250/1250
Then, I try to download some file from Internet using this proxy and the speed doesn't change, I always download at max speed.
Any idea?
Is the traffic labeled *_HIT, *_REFRESH_* or *_MISS ?
The delay pool only operates on upstream inbound server->squid traffic.
With caching and revalidation (HIT, REFRESH) a whole lot of traffic on
the squid->client connection with just 10KB/sec of inbound server traffic.
For this and several other good reasons QoS functionality in the TCP
layer is far better to use than Squid delay pools.
Amos
