There was Open vSwitch 2.2, single flow rule (ofctl) moving packets from
one physical ports to another one. Each physical port was 10Gb (dual
port Intel x520 card).
Our test setup consists of three hosts and two physical switches: a host
1 with custom traffic gen (netmap based) -> physical switch 1 -> a
host 2 with OVS -> physical switch 2 -> an end host 3
We recorded numbers on physical switches.
С уважением, | Sincerely,
Александр. | Alexander.
On 9/3/2014 4:51 PM, MCGRATH, GIL wrote:
I would be interested in what tools you are using for measurement
From: yimeng zhao <exahe...@gmail.com <mailto:exahe...@gmail.com>>
Date: Wednesday, September 3, 2014 at 8:08 AM
To: Alexander Shalimov <ashali...@arccn.ru <mailto:ashali...@arccn.ru>>
Cc: "discuss@openvswitch.org <mailto:discuss@openvswitch.org>"
<discuss@openvswitch.org <mailto:discuss@openvswitch.org>>
Subject: Re: [ovs-discuss] pps throughput depending on the number of flows
Could you specify your configuration of measurement?
2014-09-02 22:54 GMT+02:00 Alexander Shalimov <ashali...@arccn.ru
<mailto:ashali...@arccn.ru>>:
Hi,
We recently finished with throughput measurements of OVS. We saw
that phy-to-phy throughput became higher with the increasing
number of flows:
#1. 1 flow, phy-to-phy, 64 bytes packets => 1Mpps
#2. 32 flows (1 src ip, 32 dst ips), phy-to-phy, 64 bytes packets
=> 9Mpps
Could anyone explain why that happens: for 1 flow we have only
1Mpps, but for 32flows we have 9Mpps? It seems to be related to
multicore and concurrency. But where is this code located (flow
scheduling and etc)?
Thanks and Regards!
--
С уважением, | Sincerely,
Александр. | Alexander.
_______________________________________________
discuss mailing list
discuss@openvswitch.org <mailto:discuss@openvswitch.org>
http://openvswitch.org/mailman/listinfo/discuss
_______________________________________________
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
