On Wed, 15 Mar 2000, Jonathan Fortin wrote:
>
> Hello,
>
> After reading Squid Faq section 19.8 Delay Pools over and over, I
> assumed in my own brain that theyres no way to delay for my own
> thinking for this way.
>
> acl users 10.0.0.2-10.0.0.7/255.255.255.0
> acl admin 10.0.0.8-10.0.0.20/255.255.255.0
> delay pools 1
> delay_class 1 3
> delay_access 1 allow users
> delay_access 1 deny admin
> delay_parameters 1 50000/50000 -1/-1 10000/50000
>
> Meaning allocating 10kb/sec per ip in a 50kb/sec pool.
>
> That example does not work and i was wondering if someone can help me out.
Your "users" and "admin" acl syntax is incorrect -- you're missing the
'src' type. Denying the admin networks is redundant, since they
aren't allowed into the pool in the first place. Even if they were,
you would want to deny them first, and then allow the others. Also,
delay_class 2 would work better than 3 in this case, since you've
basically done the same thing by allowing unlimited per-network
bandwidth.
Try this instead:
acl users src 10.0.0.2-10.0.0.7/255.255.255.0
delay_pools 1
delay_class 1 2
delay_parameters 1 50000/50000 10000/50000
delay_access allow users
Keep in mind that just 5 users using 100% of their bandwidth will be
able to clog everything up for everybody else since you've put a
50000byte/sec aggregate limit on things. I recommend either -1/-1 for
aggregate, or lowering the per-machine limits even further if you
think that will become a problem.
I'm doing here almost exactly what you are attempting to do. The only
difference is that by default I am not putting aggregate or
per-network limits on bandwidth (since the network will do that by
itself), just per-machine limits. I can attest that it is working
quite well. :-)
-- Chris Dillon - cdillon@wolves.k12.mo.us - cdillon@inter-linc.net
FreeBSD: The fastest and most stable server OS on the planet.
For Intel x86 and Alpha architectures. ( http://www.freebsd.org )
Received on Wed Mar 15 2000 - 13:51:14 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:52:15 MST