Gentoo Archives: gentoo-user

From: Richard Fish <bigfish@××××××××××.org>
To: gentoo-user@l.g.o
Subject: Re: [gentoo-user] SATA tuning ?
Date: Wed, 16 Aug 2006 04:51:34
Message-Id: 7573e9640608152147l4497bd11o6b3a45bd746ca114@mail.gmail.com
1 Moving this back to gentoo-user, as I accidentally replied off list.
2
3 Meino, please don't CC me directly on replies. I'll read them on the list...
4
5 On 8/15/06, Meino Christian Cramer <Meino.Cramer@×××.de> wrote:
6 > > > WHen doing things, which mixes higher CPU-loads with massive hd
7 > > > utilization, things are going slow (compilation of Blender for
8 > > > example).
9 > >
10 > > Ok, let's try to test that. We'll start by saturating your CPU(s).
11 > > On one terminal start "bzip2 -9 < /dev/urandom >/dev/null".
12 >
13 > This commandline puts a BIG SMILE onto my face ! Yes, this is as
14 > simple as it is genious!!! Great! Really a nice CPU barbeque !
15 >
16 > > (If you
17 > > have multiple processors, start one of these bzip2 commands on one
18 > > terminal for each processor you have).
19 > >
20 > > Then on another, repeat the "hdparm -Tt /dev/sda"
21 >
22 > These are the results __without__ the CPU roaster:
23 >
24 > solfire:Mail/vim>sudo hdparm -tT /dev/sda
25 > /dev/sda:
26 > Timing cached reads: 2996 MB in 2.00 seconds = 1499.13 MB/sec
27 > Timing buffered disk reads: 174 MB in 3.01 seconds = 57.79 MB/sec
28 >
29 > and this are the results __with__ the CPU roaster:
30 >
31 > solfire:/home/mccramer>sudo hdparm -Tt /dev/sda
32 > solfire:/home/mccramer>sudo hdparm -Tt /dev/sda
33 > /dev/sda:
34 > Timing cached reads: 2160 MB in 2.10 seconds = 1030.12 MB/sec
35 > Timing buffered disk reads: 174 MB in 3.03 seconds = 57.41 MB/sec
36 >
37 >
38 > The chached reads dropped by ~469MByte/s. The buffered reads are
39 > nearly the same.
40
41 The buffered reads are all we care about. They are the actual reads
42 from the disk to RAM. The cached reads is just a repeated read of the
43 same sector of the disk, so today is really just a test of your memory
44 bandwidth. Since we are loading memory and the CPU pretty heavily for
45 this test, and significant drop is to be expected.
46
47 So, it is not CPU utilization that is hurting your performance.
48
49 You mentioned problems compiling. The most likely case I can think of
50 is that you do not have enough memory, and are inducing the system to
51 swap. Indeed when compiling most programs, you should see very little
52 if any disk activity. This is particularly suspect if you have
53 something like MAKEOPTS=-j4.
54
55 Regards,
56 -Richard
57 --
58 gentoo-user@g.o mailing list

Replies

Subject Author
Re: [gentoo-user] SATA tuning ? Meino Christian Cramer <Meino.Cramer@×××.de>