Re: Can't get the ingress policer to work

[Sebastian Arcus <shop@open-t.co.uk>]

I was just using an scp command. I figured as, with ingress, everything 
goes into a single queue - it should show fairly relevant results - even 
if it wasn't down to the tiniest fraction of precision.



On 09/05/13 10:51, Ian Macintosh wrote:
> How are you testing the rate?  If you're using some random protocol
> that will cloud the issue.  Try using raw tcp/udp via a test tool.
> iperf comes to mind.
>
> On 9 May 2013 09:57, Sebastian Arcus <shop@open-t.co.uk> wrote:
>> I'm trying to limit the bandwidth on the ingress leg of my Internet
>> connection. I'm using the code from lartc.org:
>>
>>
>> tc qdisc add dev eth0 handle ffff: ingress
>>
>> tc filter add dev eth0 parent ffff: protocol ip prio 50 \
>>      u32 match ip src 0.0.0.0/0 police rate 500kbit \
>>      burst 10k drop flowid :1
>>
>>
>> I've tried rates of 1Mbit, 10Mbit, 100Mbit. I've tried bursts of 1k, 10k,
>> 100k, 1000k. Nothing seems to make any difference. The test download starts
>> at about 20Mbytes/second - then keeps on slowing down and stalling
>> intermittently all the way down to 6kbytes/second. Then it bobs up and down,
>> stalling all the time, between 5kbytes/second and 17kbytes/second. There
>> seems to be absolutely no relation between the rate I set and the resulting
>> download rate.
>>
>> I'm testing on two VM's. With no traffic shaping on, I get about 36
>> megabytes/second clean speed between the two vms. Kernel on both sides is
>> 3.8.4. I'm only applying traffic shaping on one of the vm's.
>>
>> Any suggestions would be much appreciated. I ran out of ideas so far. I've
>> reread the tc-htb man page, searched google on the ingress queue - but can't
>> see what am I missing.
>> --
>> To unsubscribe from this list: send the line "unsubscribe lartc" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>
>


-- 
Linux vehicle CCTV - www.open-t.co.uk/iroko