From owner-freebsd-current Tue Sep 19 02:59:38 1995 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id CAA23114 for current-outgoing; Tue, 19 Sep 1995 02:59:38 -0700 Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id CAA23106 for ; Tue, 19 Sep 1995 02:59:32 -0700 Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.6.12/8.6.12) id CAA05609; Tue, 19 Sep 1995 02:58:40 -0700 From: "Rodney W. Grimes" Message-Id: <199509190958.CAA05609@GndRsh.aac.dev.com> Subject: Re: Which SUP files are available and where ? To: bde@zeta.org.au (Bruce Evans) Date: Tue, 19 Sep 1995 02:58:40 -0700 (PDT) Cc: bde@zeta.org.au, current@freebsd.org, rkw@dataplex.net, wollman@lcs.mit.edu In-Reply-To: <199509190718.RAA30253@godzilla.zeta.org.au> from "Bruce Evans" at Sep 19, 95 05:18:10 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2110 Sender: owner-current@freebsd.org Precedence: bulk > > >> >For those who have a high-speed net connection, CTM is a lose. It's > >> > >> This assumes an infinitely high-speed server or a small number of > >> clients. > > >You run out of ethernet before you run out of CPU power on anything > >faster than a DX4/100 with 32MB of memory doing sup services. A > >Pentium 100 can keep a 100Mb/s pipe clear full of sup with 64Mb of > >memory. I would call that very very far from ``infinitely high-speed'' > >and actually falls into the mid-range systems for me. > > I thought that the CPU ran out of power before the pipe was half full, > even doing raw data movement for nfs. I have done iozones over NFS on 100BaseTx networking and seen numbers well in excess of 3MBytes/s reading (forget about writting, we all know sync nfs is dog slow at that). > For sup it will have to traverse > file systems so it will be hard to get more than 1MB of throughput per > file system. File system bandwidth is not a problem. Again, I can produce iozone results in excess of 6MB/sec quite easily on local fast disks. >How much throughput can you get through sup? That is one I have not measures. > Is it as fast as cvs co ;-). A _LOT_ faster when you are talking about the two running over local ethernet. NFS gets in the way a bit. Sup is slow over long RTT links due to the 2 RTT needed for many of the things it does, it is blazing fast on local networks (and smokes on 100Mb/s networks :-)). I can probably sup a src tree via localhost faster than I could cvs co the same one if you wanted to test this out on a local machine. cvs co is slow due to all the . lock files that go on, sup is fast as it reads an index file and starts pumping data. > When did FreeFall get a 100 Mb/s pipe? It didn't, I have 100BaseTX here, hub and all.... NFS diskless is about to become real for me again :-) And the server is on a UPS with very realiable hardware so my writes are going to be async and fast :-). -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD