From mboxrd@z Thu Jan 1 00:00:00 1970 From: Marcelo Ricardo Leitner Subject: Re: GRO aggregation Date: Tue, 11 Sep 2012 15:48:13 -0300 Message-ID: <504F876D.9020102@redhat.com> References: <504F4063.9030706@mellanox.com>,<504F80DC.7010707@redhat.com> <36F7E4A28C18BE4DB7C86058E7B607241E622015@MTRDAG01.mtl.com> Reply-To: mleitner@redhat.com Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: "netdev@vger.kernel.org" To: Shlomo Pongratz Return-path: Received: from mx1.redhat.com ([209.132.183.28]:54323 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751895Ab2IKSsT (ORCPT ); Tue, 11 Sep 2012 14:48:19 -0400 In-Reply-To: <36F7E4A28C18BE4DB7C86058E7B607241E622015@MTRDAG01.mtl.com> Sender: netdev-owner@vger.kernel.org List-ID: On 09/11/2012 03:41 PM, Shlomo Pongratz wrote: > From: Marcelo Ricardo Leitner [mleitner@redhat.com] > Sent: Tuesday, September 11, 2012 9:20 PM > To: Shlomo Pongratz > Cc: netdev@vger.kernel.org > Subject: Re: GRO aggregation > > On 09/11/2012 10:45 AM, Shlomo Pongartz wrote: >> Hi, >> >> I=92m checking GRO aggregation with kernel 3.6.0-rc1+ using Intel ix= gbe >> driver. >> The mtu is 1500 and GRO is on and so are SG and RX checksum. >> I ran iperf with default setting and monitor the receiver with tcpdu= mp. >> The tcpdump shows that the maximal aggregation is 32120 which is 21 = * 1500. >> In the transmitter side tcpdump shows that TSO works better (~64K). >> I did a capture without GRO enabled to see if there was a difference >> between any flag >> of any two consecutive packets that forced flushing but didn't find >> anything. >> Is the GRO aggregation can be tuned. > > Hi Shlomo, > > Have you tried tuning coalescing parameters? > > Marcelo > > > Hi Marcelo > > I didn't play with interrupts coalescing. > Do you suggest to increase the value? Actually it was an idea from top of my mind, I don't know how it applie= s=20 to ixgbe, sorry. But making the NIC hold the packets a bit more should=20 make it send larger ones to kernel. Trade-off between latency/throughpu= t. I was thinking about ethtool -c options, like rx-usecs* Marcelo