Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 10 Apr 2007 03:45:53 -0700
From:      soralx@cydem.org
To:        freebsd-hackers@freebsd.org
Cc:        youshi10@u.washington.edu
Subject:   Re: Discovering list of open files from "kernel level" without using utils like lsof
Message-ID:  <20070410034553.3afa9ea9@soralx.cydem.org>
In-Reply-To: <Pine.LNX.4.43.0704091149410.23549@hymn09.u.washington.edu>
References:  <20070409152956.GB93087@dan.emsphone.com> <Pine.LNX.4.43.0704091149410.23549@hymn09.u.washington.edu>

next in thread | previous in thread | raw e-mail | index | archive | help

good_localtime(all);

> I'm thinking of changing "spinning down" to "adjusting acoustics
> level", i.e. reduce spindle speed to avoid pricey spin-downs, which
> may cause long-term serious damage to disks.

Whoa, hold your horses.
Recall that the friction between air and rotating plates inside your
winchester creates a definite gas flow. I can think of at least 3
design objectives (there could be more) that utilize that circulation:

    0a. Float the magnetic head micrometers above plate's surface;
        AFAICT, it is this fluid 'cushion' that resists any force
        directed normal to the head towards the plate, thus
        preventing head crashes.
    0b. Circulate the inside air throug a filter and a dessicant.
    0c. Detract arm's stopper when the plates get up to speed.
    0d. Cooling of other components?

Modern low-speed (7200 RPM) IDE hard drives have bearings quiet enough
to cause little trouble. You're better off switching the 'acoustic
mode' setting (or what ever it is called) from 'performace' mode to
'low noise'. I know, it's hard (lowering performace of the 'puter with
your own hands -- OMG!), and the heart rightfully rebels agains such a
[lacking the right word], but mediate, let out your deep Zen powers, and
just do it ;) This way, the drive should try to keep itself quiet
(not throw the arm round too fast, slow down [maybe?] when possible),
and it should really know better how and when to do that (given that the
factory blessed it's brain with a good firmware). 'Tis supposed to be
better than just slowing down the motor.

> Trying to keep my electric bills down and keep my disk life up :).

00. Risking disk life to save a few watts? That's rather crazy, or too
    much courage? Undue stress on electronics (the inverter, plus maybe
    something else) as well...
01. I know that some gyro rotors (which use ball bearing and spin at
    22k RPM) may last a decade or even two. If you don't quite believe
    me, then you're right: they are not running with 100% duty cycle.
    Still, the idea here is that precision ball bearings may last for
    some time. Now, sleeve bearings (such as those used in modern hard
    drives  -- the fancy term 'hydrodynamic uber-duper whatever" means
    not a lot more than that) last even longer. Sometimes far longer.
    I've seen a lot of failed disks in my life (notably, all of them
    IDE), but none died because of bearing failure. In fact, I still
    have an ancient Seagate Barracuda that sounds like a whole platoon
    of high speed drills working their chucks off, but still is
    functional. BTW, I've done autopsy on many of them, and the most
    common causes of malfunction are:
	0a. electronics going crazy in the head;
	0b. head crash.

> -Garrett

[SorAlx]  ridin' VN1500-B2



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070410034553.3afa9ea9>