Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 25 Aug 2016 10:55:30 -0700
From:      hiren panchasara <hiren@strugglingcoder.info>
To:        Andrew Gallatin <gallatin@FreeBSD.org>
Cc:        src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-head@freebsd.org
Subject:   Re: svn commit: r303626 - in head/sys: netinet netinet6
Message-ID:  <20160825175530.GE41576@strugglingcoder.info>
In-Reply-To: <201608011702.u71H2L1e095815@repo.freebsd.org>
References:  <201608011702.u71H2L1e095815@repo.freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help

--WK3l2KTTmXPVedZ6
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On 08/01/16 at 05:02P, Andrew Gallatin wrote:
> Author: gallatin
> Date: Mon Aug  1 17:02:21 2016
> New Revision: 303626
> URL: https://svnweb.freebsd.org/changeset/base/303626
>=20
> Log:
>   Rework IPV6 TCP path MTU discovery to match IPv4
>  =20
>   - Re-write tcp_ctlinput6() to closely mimic the IPv4 tcp_ctlinput()
>  =20
>   - Now that tcp_ctlinput6() updates t_maxseg, we can allow ip6_output()
>     to send TCP packets without looking at the tcp host cache for every
>     single transmit.
>  =20
>   - Make the icmp6 code mimic the IPv4 code & avoid returning
>     PRC_HOSTDEAD because it is so expensive.
>  =20
>   Without these changes in place, every TCP6 pmtu discovery or host
>   unreachable ICMP resulted in a call to in6_pcbnotify() which walks the
>   tcbinfo table with the write lock held.  Because the tcbinfo table is
>   shared between IPv4 and IPv6, this causes huge scalabilty issues on
>   servers with lots of (~100K) TCP connections, to the point where even
>   a small percent of IPv6 traffic had a disproportionate impact on
>   overall throughput.
>  =20
>   Reviewed by:	bz, rrs, ae (all earlier versions), lstewart (in Netflix's=
 tree)
>   Sponsored by:		Netflix
>   Differential Revision:	https://reviews.freebsd.org/D7272

Drew,

What do you think about getting this into stable/11?

Cheers,
Hiren

--WK3l2KTTmXPVedZ6
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
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=
=Siao
-----END PGP SIGNATURE-----

--WK3l2KTTmXPVedZ6--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20160825175530.GE41576>