From owner-freebsd-current Thu Nov 13 09:06:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id JAA27220 for current-outgoing; Thu, 13 Nov 1997 09:06:42 -0800 (PST) (envelope-from owner-freebsd-current) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id JAA27209 for ; Thu, 13 Nov 1997 09:06:36 -0800 (PST) (envelope-from fbsd@Ilsa.StevesCafe.com) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.7/8.8.5) with ESMTP id KAA01538; Thu, 13 Nov 1997 10:02:56 -0700 (MST) Message-Id: <199711131702.KAA01538@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: Thomas Dean cc: drussell@saturn-tech.com, gurney_j@resnet.uoregon.edu, bsampley@bsampley.vip.best.com, current@FreeBSD.ORG Subject: Re: make world time???/ In-reply-to: Your message of "Wed, 12 Nov 1997 16:31:38 PST." <199711130031.QAA02936@ix.netcom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 13 Nov 1997 10:02:56 -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi, > I don't understand how a simple build, 'make world > log.file' on a > dual P-133 system with 3.0-current can take 7 hours when you are > talking about 1.5 hours. Are we doing the same thing? Is there > something wrong with my machine or setup? I made no changes to the > files down-loaded, other than add network to /etc/group. Am I doing > something wrong? There was little other activity on the machine. > Fetchmail every 5 min. > > The time statements from log.file: > > -------------------------------------------------------------- > make world started on Mon Nov 10 10:20:31 PST 1997 > -------------------------------------------------------------- > ... > -------------------------------------------------------------- > make world completed on Mon Nov 10 17:35:28 PST 1997 > -------------------------------------------------------------- On my very similar hardware: GA586DX, 2 133MHz P5s, 2 older scsiII disks: -------------------------------------------------------------- make world started on Fri Aug 15 01:45:01 MDT 1997 -------------------------------------------------------------- ... -------------------------------------------------------------- make world completed on Fri Aug 15 07:16:04 MDT 1997 -------------------------------------------------------------- 19863.39s real 11352.56s user 8057.75s system for a total of 5 hours, 31 minutes. --- then same source, on my GA686DX, 2 200MHz P6s, 1 older 7200rpm disk: -------------------------------------------------------------- make world started on Fri Aug 15 23:44:00 MDT 1997 -------------------------------------------------------------- ... -------------------------------------------------------------- make world completed on Sat Aug 16 02:58:58 MDT 1997 -------------------------------------------------------------- 11698.66s real 4394.25s user 3963.56s system for a total of 3 hours, 14 minutes --- finally, on my Intel PR440FX, 2 200MHZ P6s, 2 7200rpm baracudas, partitioned: Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/sd0a 63503 40437 17986 69% / /dev/sd0s1h 1203775 459399 648074 41% /usr /dev/sd0s1e 47599 1164 42628 3% /var /dev/sd0f 349343 239201 82195 74% /usr/src /dev/sd1g 190543 141891 33409 81% /usr/obj -------------------------------------------------------------- make world started on Thu Aug 14 21:32:37 MDT 1997 -------------------------------------------------------------- ... -------------------------------------------------------------- make world completed on Thu Aug 14 23:58:00 MDT 1997 -------------------------------------------------------------- 8723.42s real 4387.05s user 4031.71s system for a total of 2 hours, 25 minutes --- My guess is that your disk IO is the main diff between your 7hours and my 5hours for similar hardware. Big wins are: -pipe /dev/sd0f on /usr/src (local, noatime) ^^^^^^^ /dev/sd1g on /usr/obj (asynchronous, local, noatime) ^^^^^^^^^^^^ ^^^^^^^ I believe all the above times are pre-parallelmake of the world. Although somewhat old, they are the last set I have where all 3 systems built the world at the same time. There is a big gain when going from P5 to P6, about 2:1 -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD