Date: Tue, 9 Mar 2004 08:50:58 +0100 (MET) From: Helge Oldach <helge.oldach@atosorigin.com> To: jbarrett@amduat.net (Jacob S. Barrett) Cc: julian@elischer.org Subject: Re: Solution for Resilient VLAN Trunk Bonding Message-ID: <200403090750.IAA21292@galaxy.hbg.de.ao-srv.com> In-Reply-To: <200403081553.58502.jbarrett@amduat.net> from "Jacob S. Barrett" at "Mar 9, 2004 0:53:58 am"
index | next in thread | previous in thread | raw e-mail
Jacob S. Barrett: >I have some questions about the ng_fec. Would it work if each interface >was connected to a different switch? I'd say this isn't an issue with ng_fec, but rather an architectural point regarding EtherChannel as such. I am not aware of any switch vendor that offers multi-chassis EtherChannel. In fact many even require that the physical links terminate on the same switch blade and don't permit distribution over multiple blades in the same chassis. >Everything I have read on the list says that they done it only with >having "trunking" enabled on the switch as well. That is definitely the case. Both ends must be aware that they belong to a channel, and if you want decent resiliency they should also talk the appropriate channeling protocol. Usually LACP (802.3ad); in the Cisco case PAgP might suit you better. Neither is supported by ng_fec, AFAIK. Be aware that the term "trunking" is commonly used for grouping multiple VLANs onto one link (802.1q) in the switching world. Helgehelp
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200403090750.IAA21292>