Date: Sat, 09 Jun 2001 10:52:34 -0700 From: Terry Lambert <tlambert2@mindspring.com> To: Ruslan Ermilov <ru@FreeBSD.ORG> Cc: Mike Silbersack <silby@silby.com>, freebsd-net@FreeBSD.ORG, freebsd-arch@FreeBSD.ORG Subject: Re: New TCP sequence number generation algorithm; review needed Message-ID: <3B226262.AF4858E2@mindspring.com> References: <20010608005234.W92206-200000@achilles.silby.com> <3B20A02E.B1507A80@mindspring.com> <20010609132227.B87114@sunbay.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Ruslan Ermilov wrote: > > On Fri, Jun 08, 2001 at 02:51:42AM -0700, Terry Lambert wrote: > [...] > > > > 6) This adds per-connection state, which is evil > > when you want a lot of connections: the way to > > get a lot of connections is to remove as much > > per-connection state as possible, which in turn > > reduces your per-connection resource costs, and > > that in turn increases the number of connections > > you are capable of sustaining with a constant > > set of resources. > > > Umm, this adds a per-destination, not per-connection state. For things like web servers, where the destination is a client PC with a single user, the net effect is per-connection state, since magnitude(connection) == magnitude(destination). The only escape is an edge proxy for multiple clients at a single site. I agree that this is a possibility, but I think the majority of web site traffic is single client, no pedge proxy. Note that even if there _is_ an edge proxy, that the proxy itself will satisfy the request: the origin server will never see it. As a result, we are back to ther being a direct correlation,. i.e.: magnitude(connection) == magnitude(destination), from the point of view of the server. -- Terry To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3B226262.AF4858E2>