From owner-freebsd-current Tue Nov 4 01:06:44 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id BAA26158 for current-outgoing; Tue, 4 Nov 1997 01:06:44 -0800 (PST) (envelope-from owner-freebsd-current) Received: from sendero-ppp.i-connect.net (sendero-ppp.i-Connect.Net [206.190.143.100]) by hub.freebsd.org (8.8.7/8.8.7) with SMTP id BAA26117 for ; Tue, 4 Nov 1997 01:06:34 -0800 (PST) (envelope-from shimon@sendero-ppp.i-connect.net) Received: (qmail 18989 invoked by uid 1000); 4 Nov 1997 09:07:03 -0000 Message-ID: X-Mailer: XFMail 1.2-beta-103097 [p0] on FreeBSD X-Priority: 3 (Normal) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit MIME-Version: 1.0 In-Reply-To: <199711040754.CAA00612@dyson.iquest.net> Date: Tue, 04 Nov 1997 01:07:03 -0800 (PST) Organization: Atlas Telecom From: Simon Shapiro To: dyson@FreeBSD.ORG Subject: Re: More on fast make world... Cc: freebsd-current@FreeBSD.ORG Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi John S. Dyson; On 04-Nov-97 you wrote: > Simon Shapiro said: > > I really am not so sure what takes the time, but it is not disk I/O. > > > > I setup a test machine with 128MB of RAM, a RAID-1 root disk, a RAID-0 > > 8x32KB stripes wide for /usr/src and /usr/obj, on a DPT PM3334UDW > > (Ultra, > > wide, differential). Disks are 4GB Barcudas all around. > > > > This configuration is capable of 980+ disk I/O per second on the RAID-1 > > and > > 1740+ on the RAID-0 array. > > > > Starting with a fresh install, SMP kernel current for today, DPT > > configured > > with no options, but the performance monitors and a 1 sec timeout hack > > to > > catch lost interrupts (new firmware that may be a bit buggy). > > > > Top reports (abbreviated): > > > > load averages: 7.36, 6.59, 5.06 > > CPU states: 54.1% user, 0.0% nice, 44.4% system, 1.6% interrupt, > > 0.0%idle > > Mem: 11M Active, 17M Inact, 23M Wired, 47M Cache, 8248K Buf, 26M Free > > > > Iostat says: > > > > tty sd0 sd16 cpu > > tin tout sps tps msps sps tps msps us ni sy in id > > 0 1905 771 74 0.0 668 56 0.0 27 0 51 2 20 > > > > This is typical over the last hour or so. Anything else I should try? > > > Try increasing the size of the vnode cache: > sysctl -w debug.wantfreevnodes=15000 > > Make sure you are using -pipe, if not, then use MFS for /tmp or > wherever your tempfiles go... Silly question: How do I do that? Last I tried I ended in a panic... How large? Where to mount? /tmp or /var/tmp? I foreget ... Without these improvements, we are at 102 minutes end-to-end. Let's go for sub-hour make world! > > -- > John > dyson@freebsd.org > jdyson@nc.com --- If Microsoft Built Cars: Every time they repainted the lines on the road, you'd have to buy a new car. Sincerely Yours, Simon Shapiro Atlas Telecom Senior Architect 14355 SW Allen Blvd., Suite 130 Beaverton OR 97005 Shimon@i-Connect.Net Voice: 503.799.2313