From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from thsbbfxrt02p.thalesgroup.com (thsbbfxrt02p.thalesgroup.com [192.93.158.29]) by dpdk.org (Postfix) with ESMTP id 3711B1041 for ; Tue, 12 Sep 2017 17:40:50 +0200 (CEST) Received: from thsbbfxrt02p.thalesgroup.com (localhost [127.0.0.1]) by localhost (Postfix) with SMTP id 3xs8C60dWpzG7J for ; Tue, 12 Sep 2017 17:40:50 +0200 (CEST) From: DESBRUS Maxime To: "users@dpdk.org" Date: Tue, 12 Sep 2017 17:40:48 +0200 Thread-Topic: Small constant frame loss only for more than 5 seconds of traffic at 10Gbps Thread-Index: AdMr3LqRY/IxqgtmT5qkRGSlmFdBIg== Message-ID: <6C26BDBD2FD19B418E20F4A951DDC74A1A09061FCA@THSONEA01CMS08P.one.grp> Accept-Language: fr-FR Content-Language: fr-FR X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: fr-FR x-pmwin-version: 3.1.3.0, Antivirus-Engine: 3.69.2, Antivirus-Data: 5.43 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable X-Content-Filtered-By: Mailman/MimeDel 2.1.15 Subject: [dpdk-users] Small constant frame loss only for more than 5 seconds of traffic at 10Gbps X-BeenThere: users@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list List-Id: DPDK usage discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 12 Sep 2017 15:40:50 -0000 Hello I am trying to capture 60 B Ethernet frames at link speed on two 10GbE port= s with DPDK 17.08 on Linux 4.4 and an Intel XL710 card. I use a slightly modified version of the l2fwd example program, that only c= ounts incoming frames and immediately releases the mbufs. I see a strange behavior because I can capture without any loss for 5s of t= raffic, but for more than that I always lose a few frames (about 15 000). F= or 10, 30 or 60 seconds of traffic the number of lost frames does not chang= e much (it is a constant, and not a percentage of the incoming traffic). I did apply all the recommendations to eliminate perturbations on the worki= ng cores: * isolate cores with isolcpus, nohz_full, rcu_nocbs boot parameters * disable hyper threading * disable NMI watchdog * disable pstate frequency scaling * disable pause frames on both sides I also tried to tweak a few things: * increase/decrease mbuf pool size * use a dedicated rte_mempool object per port * increase/decrease packet burst length * increase/decrease rx descriptor count Although some of these did decrease loss a bit, I still get a few thousands= lost frames for more than 5 seconds of incoming traffic. I don't understand why performance seem to change only after a few seconds = of traffic. Any ideas? Thanks -- Maxime DESBRUS