From owner-freebsd-net Wed Jun 21 7:48:53 2000 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id D7DE337B51B for ; Wed, 21 Jun 2000 07:48:49 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id KAA39286; Wed, 21 Jun 2000 10:48:46 -0400 (EDT) (envelope-from wollman) Date: Wed, 21 Jun 2000 10:48:46 -0400 (EDT) From: Garrett Wollman Message-Id: <200006211448.KAA39286@khavrinen.lcs.mit.edu> To: "C. Stephen Gunn" Cc: net@FreeBSD.ORG Subject: Re: bridge + VLAN + netgraph In-Reply-To: <20000620233346.B1128@waterspout.com> References: <200006152006.QAA06392@cholla.INRS-Telecom.UQuebec.CA> <394FAB91.794BDF32@elischer.org> <20000620233346.B1128@waterspout.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > - Since VLANs shove another 4-byte field in the Ethernet Frame, > we bump the MTU by 4 bytes. Some cards/drivers reject frames > larger than 1500 bytes as giants. This limits the interfaces > recommended for user with VLANs. As someone pointed out > in discussions with me, MTU is a TCP/IP concept, not an > intrinsic property of an interface. This paragraph is somewhat incorrect. Ethernet (that is, IEEE 802.3) defines a maximum packet length, which is 1518 octets, including header and CRC. In conjunction with IEEE 802.1p, this maximum packet length is increased to 1522 octets. Whether or not you use 802.1p/Q tagging, the MTU of an Ethernet is and remains 1500 octets; the increased maximum length of an 802.1p packet is entirely in the header. (If the MTU is not 1500 octets, it's not Ethernet.) Some network interfaces make it possible to directly configure the maximum packet length. Others consider 1522-octet frames to be errors, so the only way to enable reception of tagged frames on these interfaces is to accept error frames and then rely on additional software mechanism to distinguish long frames from those with other sorts of errors. (Typically on these types of interfaces it is necessary to manually trim off the CRC as well.) Still others are simply unable to participate in 802.1p/Q networks. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message