Date: Fri, 14 Jun 1996 10:10:24 -0400 From: Garrett Wollman <wollman@lcs.mit.edu> To: "Gary Palmer" <gpalmer@freebsd.org> Cc: bill@twwells.com (T. William Wells), freebsd-questions@freebsd.org Subject: Re: performance effects of many interface addresses? Message-ID: <9606141410.AA17238@halloran-eldar.lcs.mit.edu> In-Reply-To: <7122.834716598@palmer.demon.co.uk> References: <4pqeud$ohs@twwells.com> <7122.834716598@palmer.demon.co.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
<<On Fri, 14 Jun 1996 02:43:18 +0100, "Gary Palmer" <gpalmer@freebsd.org> said: > he got bored! Also, the overhead only counts when you have to search > the alias list. If you are doing transfers through the main IP > address, the transfer rate shouldn't be affected (if I remember right) Wrong on two counts: 1) There is no such thing as an `alias list'. There is an `address list'. 2) For every packet received, the system much search (potentially) the entire address list (potentially twice) to determine whether it should accept or forward/drop the packet. In 2.1 and -stable, the system must in addition search the entire address list to determine whether packets being sent are broadcasts or not. (Structurally, it would have been far better for the API to be more like the multicast API, where you use the same address and a special setsockopt() to set the outgoing interface.) -GAWollman -- Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. Opinions not those of| It is a bond more powerful than absence. We like people MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9606141410.AA17238>