Re: [squid-users] peer weight is not working please help

From: Chris Robertson <crobertson@dont-contact.us>
Date: Tue, 20 Jun 2006 10:37:08 -0800

squid learner wrote:

>a little more with advance THANKS
>here you see the access log the peer sellection is
>very fast
>and it goes to other with out completing req
>
>
This is working as designed. Each HTTP request is (should be) a stand
alone request. No HTTP request relies on any other to finish.

> is there any way to reduce this
>that every peer should change after some time like
>after 5 minuts or so
>you see the acces log it changes very fast
>and when it change it leave the req incomplete
>because every peer is given direction to diffrent
>router
>
>
This is incorrect. The request is being passed to the parent, but the
parent is not responding in a timely fashion (over six seconds in one
instance). The user controlling the client gets tired of waiting, and
hits "stop" or "reload" on the browser, which causes Squid to log a
TCP_MISS/000. More interpretation follows...

>Kernel IP routing table
>Destination Gateway Genmask Flags
>Metric Ref Use Iface
>212.107.116.243 192.168.1.2 255.255.255.255 UGH
>0 0 0 eth0
>212.162.158.85 192.168.1.3 255.255.255.255 UGH
>0 0 0 eth0
>212.92.193.87 192.168.1.1 255.255.255.255 UGH
>0 0 0 eth0
>192.168.1.0 * 255.255.255.0 U
>0 0 0 eth0
>169.254.0.0 * 255.255.0.0 U
>0 0 0 eth0
>127.0.0.0 * 255.0.0.0 U
>0 0 0 lo
>default 192.168.1.3 0.0.0.0 UG
>0 0 0 eth0
>default 192.168.1.2 0.0.0.0 UG
>0 0 0 eth0
>default 192.168.1.1 0.0.0.0 UG
>0 0 0 eth0
>[root@modi root]#
>
>
>
>1150813218.959 556 192.168.1.23 TCP_MISS/204 258
>GET http://www.google.com/url? -
>ROUNDROBIN_PARENT/proxy.jeel.com text/html
>
>
This cache parent was successfully contacted. Response contained no content

>1150813225.025 6494 192.168.1.23 TCP_MISS/000 0 GET
>http://www.newscientist.com/ -
>ROUNDROBIN_PARENT/proxy.cyberia.net.sa -
>
>
This request was canceled by the client before any data was returned.

>1150813225.947 902 192.168.1.23 TCP_MISS/302 567
>GET http://www.newscientist.com/ - NONE/- text/html
>
>
No parent was used.

>1150813227.375 1428 192.168.1.23 TCP_MISS/000 0 GET
>http://www.newscientist.com/home.ns -
>ROUNDROBIN_PARENT/proxy.cyberia.net.sa -
>
>
This request was canceled by the client before any data was returned.

>1150813228.705 1424 192.168.1.23 TCP_MISS/000 0 GET
>http://www.google.com/url? -
>ROUNDROBIN_PARENT/proxy.awalnet.net.sa -
>
>
This request was canceled...

>1150813228.713 3858 192.168.1.23 TCP_MISS/000 0 GET
>http://www.google.com/url? -
>ROUNDROBIN_PARENT/proxy.awalnet.net.sa -
>
>
This request was canceled...

>1150813229.625 2204 192.168.1.23 TCP_MISS/200 28299
>GET http://content.nejm.org/ - NONE/- text/html
>
>
Successful request. No parent used.

>1150813229.629 905 192.168.1.23 TCP_MISS/000 0 GET
>http://content.nejm.org/homepage.css -
>ROUNDROBIN_PARENT/proxy.cyberia.net.sa -
>
>
This request was canceled...

I don't see a single successful request made through a parent cache.
You need to figure out why the parent caches are not passing you any
data. Round robin will bounce every request to a different parent,
true, but it will not cause a request to canceled before it is serviced.

Chris
Received on Tue Jun 20 2006 - 12:37:39 MDT

This archive was generated by hypermail pre-2.1.9 : Sat Jul 01 2006 - 12:00:02 MDT