Date: Thu, 5 Jul 2012 12:00:07 +0300 From: Vyacheslav Kulikovskyy <coolsysop@gmail.com> To: Eugene Grosbein <egrosbein@rdtc.ru> Cc: freebsd-net@freebsd.org Subject: Re: lagg speed trouble Message-ID: <CABj7HZ_2q70FGFOD_QB7ATVpZ68n0ZQroi0OV8xkpSxajqeE1A@mail.gmail.com> In-Reply-To: <4FF5329E.1000007@rdtc.ru> References: <CABj7HZ8dnbpwhgpQ9pE=M2xJh_LFp2oYdPqvXke0GF%2B4rt7sgg@mail.gmail.com> <4FF5329E.1000007@rdtc.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
2012/7/5 Eugene Grosbein <egrosbein@rdtc.ru>
> 04.07.2012 18:30, Vyacheslav Kulikovskyy =D0=C9=DB=C5=D4:
> > i have sever with two 1G links (em) aggregated by lagg0
> >
> > after 1700Megabits i have collisions/errors on lagg0 port, but not on e=
m0
> > or em1
> >
> > I'm using nginx in own CDN. and server don't limited my mbufs, irq, or
> > anything else.. only lagg0 errors (
>
> lagg0 by default uses way too small output FIFO queue, 50 slots.
> You definitely want to increase this to sum of your em intefaces
> txd buffers at least. So, if you use hw.em.txd=3D4096, you should
> use 8192 for lagg or more. I prefer 10K, in /boot/loader.conf:
>
> net.link.ifqmaxlen=3D10240
>
> Reboot is required.
>
> Eugene Grosbein
>
This is solution don't help me
hw.em.txd: 4096
net.link.ifqmaxlen: 10240
# netstat -w 1 -I lagg0
input (lagg0) output
packets errs idrops bytes packets errs bytes colls
77600 0 0 4771508 71175 70 208916199 0
79747 0 0 4885070 73101 242 218847447 0
84956 0 0 5262540 74504 455 213752515 0
81333 0 0 4998499 73836 229 215808086 0
and no errors on em0,em1
more info: FreeBSD 9.0-STABLE #0: AMD64
ps after small test i'll try patch from Andrew Thompson
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CABj7HZ_2q70FGFOD_QB7ATVpZ68n0ZQroi0OV8xkpSxajqeE1A>