[ovs-discuss] QoS Rate limiting example for physical interfaces - reg
cerisierfilix at hotmail.com
Thu Dec 27 04:03:05 UTC 2012
No. I have not tried passing traffic between two machines. I will try it. But can't we use iperf to pass traffic between two IPs on the same machine just to test it?
Date: Wed, 26 Dec 2012 08:35:16 -0600
Subject: Re: [ovs-discuss] QoS Rate limiting example for physical interfaces - reg
From: brad.hedlund at gmail.com
To: cerisierfilix at hotmail.com
CC: discuss at openvswitch.org
Your iPerf session is between two IPs on the same machine. Have you tried passing traffic between two different machines?
On Dec 26, 2012 3:48 AM, "Mathumitha S" <cerisierfilix at hotmail.com> wrote:
I have a quad-port in my system of which I have connected interfaces eth5 and eth4 as OvS ports 1 and 2 respectively. And I used local IPs for the two interfaces i.i I created the IPs using ifconfig command.(eth4-10.1.6.1 and eth5-10.1.6.2)
I used the following commands:
ovs-vsctl set Interface eth4 ingress_policing_rate=1000
ovs-vsctl set Interface eth4 ingress_policing_burst=100
ovs-vsctl set Interface eth5 ingress_policing_rate=10000
ovs-vsctl set Interface eth5 ingress_policing_burst=1000
I used the following iperf commands to generate traffic:
iperf -s -B 10.1.6.1
iperf -c 10.1.6.1 -B 10.1.6.2
But there is no rate-limiting done. The BW of the traffic transfer is about 40 Gbits/sec.
Also, I tried to create queues on eth5 and enqueue traffic from eth4 to a queue in eth5.
The commands I used are:
ovs-vsctl set port eth5 qos=@newqos -- --id=@newqos create qos type=linux-htb other-config:max-rate=20000 queues=0=@q0,1=@q1 -- --id=@q0 create queue other-config:min-rate=10000 other-config:max-rate=10000 -- --id=@q1 create queue other-config:min-rate=5000 other-config:max-rate=5000
ovs-ofctl mod-flows br0 "in_port=2,dl_type=0x0800,nw_proto=6,tp_dst=6000 actions=enqueue:1:1"
There is no limiting of the rate in this case too.
Can someone tell me where I am going wrong?
discuss mailing list
discuss at openvswitch.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the discuss