Re: [squid-users] Cache peer does not work

From: Amos Jeffries <squid3_at_treenet.co.nz>
Date: Wed, 25 May 2011 21:19:10 +1200

On 16/05/11 18:35, Dr. Muhammad Masroor Ali wrote:
> It was not working when I used
> cache_peer 172.16.101.3 parent 3128 3130 default
>
>
> Using 0 place of 3130 solved the problem. So, ICP port was not working.
>
> However, could somebody please explain this.
>
> It does not work when the line is,
>
> cache_peer 172.16.101.3 parent 3128 3130 default
>
> So, I change it to
> cache_peer 172.16.101.3 parent 3128 0 default
>
> It works.
>
> But is keeps working, even after reload or restart, when the line is
> changed back to
>
> cache_peer 172.16.101.3 parent 3128 3130 default
>
> I am trying to find out the reason behind this without much success.
>

It takes a reasonably large number of failed checks (10 in a row) to
declare a peer DEAD. It may eventually reach that state again someday
and repeat the problem. Until then HTTP requests will be tried to it as
well as ICP, if either ICP or ICP succeeds the check-failed count is
reset to zero.

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE9 or 3.1.12
   Beta testers wanted for 3.2.0.7 and 3.1.12.1
Received on Wed May 25 2011 - 09:19:17 MDT

This archive was generated by hypermail 2.2.0 : Wed May 25 2011 - 12:00:03 MDT