From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: IGMP Join dropping multicast packets Date: Mon, 16 Mar 2009 20:01:01 +0100 Message-ID: <49BEA1ED.4010907@cosmosbay.com> References: <91bdcedb0903141316j2dbf4160wb348a5a9e3bde8ad@mail.gmail.com> <49BC69D5.5000002@cosmosbay.com> <91bdcedb0903151904x1066ac24h63557b588e7c4967@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: netdev@vger.kernel.org To: Dave Boutcher Return-path: Received: from gw1.cosmosbay.com ([212.99.114.194]:54383 "EHLO gw1.cosmosbay.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758571AbZCPTBJ convert rfc822-to-8bit (ORCPT ); Mon, 16 Mar 2009 15:01:09 -0400 In-Reply-To: <91bdcedb0903151904x1066ac24h63557b588e7c4967@mail.gmail.com> Sender: netdev-owner@vger.kernel.org List-ID: Dave Boutcher a =E9crit : > On Sat, Mar 14, 2009 at 9:37 PM, Eric Dumazet w= rote: >> Dave Boutcher a =E9crit : >>> I'm running into an interesting problem with joining multiple >>> multicast feeds. If you join multiple multicast feeds using >>> setsockopt(...,IP_ADD_MEMBERSHIP...) it causes packets on UNRELATED >>> multicast feeds to get dropped. We have a multicast feed on a rock >>> solid network, and we were very surprised to see dropped packets. = The >>> cause was a different process/program being run by a different user >>> joining a bunch of mulitcast feeds. >> I could not reproduce the problem on my machines (bnx2 adapter), eve= n if changing >> NUMSOCK from 55 to 200 in joiner.c >=20 > Thanks for trying Eric. Based on your email I did some more testing > and thus far I've > only recreated this on x86_64 arches, not on i386. Which arch did yo= u > try it on? I tried both, 32 and 64 bit kernels. No problems so far. Could you post a linux kernel .config of a non 'working' machine, and d= mesg output ? >=20 >> Is your network a 100Mb one or Gigabit ? >> Try to slow down your joiner ? >> (Could be a flood of IGMP messages your router/switch cannot cope wi= th) >> >> Please describe your "rock solid" network setup (kind of network ada= pters you have, kind of router...) >=20 > The problem originally manifest itself at work on a 24-core Dell > server with 6 NICs. The network > is gigabit with a Cisco 4900 switch. I recreated it in my basement o= n > my little white-box > system and a cheap netgear switch. The NIC at work is Intel e1000e > driver, the one > at home is also e1000. >=20 >> If using tcpdump to force promiscuous mode on the device also trigge= rs packet losses ? >> >> (see also ifconfig ethX promisc|allmulti) >=20 > I haven't had a chance to play with promiscuous yet... >=20