Date: Mon, 01 Mar 1999 20:43:09 -0600 From: David Kelly <dkelly@hiwaay.net> To: "Matthew D. Fuller" <fullermd@futuresouth.com> Cc: freebsd-stable@FreeBSD.ORG Subject: Re: rc5des slows tape thruput Message-ID: <199903020243.UAA02030@nospam.hiwaay.net> In-Reply-To: Message from "Matthew D. Fuller" <fullermd@futuresouth.com> of "Sun, 28 Feb 1999 22:52:43 CST." <19990228225243.W3203@futuresouth.com>
next in thread | previous in thread | raw e-mail | index | archive | help
"Matthew D. Fuller" writes: > On Sun, Feb 28, 1999 at 09:57:44PM -0600, a little birdie told me > that David Kelly remarked > > > > But Friday night was getting half this thruput. Fired up X and repeated. > > Essentially the same results. Then fired up rc5des, which is niced and > > fell to an average of 185k/sec. Snipped from top: > > > > 294 dkelly 105 20 740K 520K RUN 9:58 97.95% 97.95% rc5des > > One thing to bear in mind is that a process nice'd to 20 still gets CPU > time allotted. I always idprio rc5des and friends. For instance, in > this case, run (as root, I think you have to) > idprio 31 -294 > > Then it'll show up as nice 52, and ONLY use idle CPU cycles, instead of > having an (albeit small) slice of cycles always allocated it. There is more to the story than I'm letting on just yet as I'm trying to keep things simple and then rc5des popped up as a quick example to demonstrate the problem. > No, I don't think it's 'good' that nice'd to 20 it slows the tape by that > much, but this is at least a partial workaround (and the better way to do > it IMO anyway). > > Have you tried running rc5des under 2.2.5 and seeing what speed you got? I used to run rc5des under 2.2.* on this machine and thought I got full tape speeds there. Have a P-133 on my desk at work with 2.2.8 that I can put a 400k/sec DDS-1 drive on, and rc5des, and see what happens. But back to the long story: What I'm trying to do is use FreeBSD as a tape duplication host. At work I have a Feb 8 3.0 system with (4) DDS-3 tape drives attached, two per SCSI bus, HD on a 3rd SCSI (it got that way because that's what cables and controllers I had laying around, The Powers That Be purchased $4k in tape drives and no cables.) When reading a DDS-3 tape and nothing else going on, "systat -v" shows the tape device averaging around 1.04M/sec. Having dd'ed an image of this tape to disk, using dd the other way to write is averaging 550k/ sec. Oh, but in this case I'm writing the image to all 4 tape drives at once. Rates reported by "systat -v" are wildly fluxuating. Thruput is 550k/sec averaged over 6G. Am thinking the situation where I'm writing the same data to 4 tape drives at once is similar to what happens when rc5des consumes all the spare CPU cycles. Writing to DDS-1 tapes, 4 at once, usually shows a stable 700k/sec for each tape drive. Sometimes one gets behind. Then the data rate off the HD climbs from 700k/sec. CPU utilization on a P-II 233 MHz is about 8% when writing all 4 tape drives. -- David Kelly N4HHE, dkelly@nospam.hiwaay.net ===================================================================== The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199903020243.UAA02030>