From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ben Greear Subject: Re: OOM when adding ipv6 route: How to make available more per-cpu memory? Date: Sat, 06 Nov 2010 10:08:44 -0700 Message-ID: <4CD58B9C.2030006@candelatech.com> References: <4CD43C87.5040403@candelatech.com> <1288980361.2882.1070.camel@edumazet-laptop> <4CD449A5.5070305@candelatech.com> <1288988403.2665.268.camel@edumazet-laptop> <1288995103.2665.653.camel@edumazet-laptop> <4CD49C2F.3060904@candelatech.com> <1289028392.2665.2418.camel@edumazet-laptop> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: NetDev , linux-kernel , Tejun Heo To: Eric Dumazet Return-path: In-Reply-To: <1289028392.2665.2418.camel@edumazet-laptop> Sender: linux-kernel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On 11/06/2010 12:26 AM, Eric Dumazet wrote: > Le vendredi 05 novembre 2010 =C3=A0 17:07 -0700, Ben Greear a =C3=A9c= rit : > >> We just tried creating 1000 macvlans with IPv6 addrs on a 64-bit mac= hine >> with 12GB RAM. Only around 520 interfaces properly set their IPs, a= nd >> again there are errors about of-of-memory from 'ip', but no obvious >> splats in dmesg. >> >> 'top' shows 10G or so free. >> >> It will take some time to figure out what exactly is returning >> the ENOMEM.... > > At least, nothing to do with percpu stuff ? At least I don't see any percpu dumps in dmesg. I vaguely remember someone posting some ipv6 address scalability patches some time back. I think they had to hack on /proc fs as well. I'll see if I can dig those up. > Make sure udev / hotplug is not the problem, if you create your devic= es > very fast. We can create the macvlans w/out problem, though I'm sure that could be sped up. The problem is when we try to add IPv6 addresses to them. Thanks, Ben --=20 Ben Greear Candela Technologies Inc http://www.candelatech.com