1 |
From: "Richard Fish" <bigfish@××××××××××.org> |
2 |
Subject: Re: [gentoo-user] SATA tuning ? |
3 |
Date: Tue, 15 Aug 2006 21:47:20 -0700 |
4 |
|
5 |
> Moving this back to gentoo-user, as I accidentally replied off list. |
6 |
> |
7 |
> Meino, please don't CC me directly on replies. I'll read them on the list... |
8 |
|
9 |
...sorry...my fault...bu the previous mail was a private one to me... |
10 |
|
11 |
> On 8/15/06, Meino Christian Cramer <Meino.Cramer@×××.de> wrote: |
12 |
> > > > WHen doing things, which mixes higher CPU-loads with massive hd |
13 |
> > > > utilization, things are going slow (compilation of Blender for |
14 |
> > > > example). |
15 |
> > > |
16 |
> > > Ok, let's try to test that. We'll start by saturating your CPU(s). |
17 |
> > > On one terminal start "bzip2 -9 < /dev/urandom >/dev/null". |
18 |
> > |
19 |
> > This commandline puts a BIG SMILE onto my face ! Yes, this is as |
20 |
> > simple as it is genious!!! Great! Really a nice CPU barbeque ! |
21 |
> > |
22 |
> > > (If you |
23 |
> > > have multiple processors, start one of these bzip2 commands on one |
24 |
> > > terminal for each processor you have). |
25 |
> > > |
26 |
> > > Then on another, repeat the "hdparm -Tt /dev/sda" |
27 |
> > |
28 |
> > These are the results __without__ the CPU roaster: |
29 |
> > |
30 |
> > solfire:Mail/vim>sudo hdparm -tT /dev/sda |
31 |
> > /dev/sda: |
32 |
> > Timing cached reads: 2996 MB in 2.00 seconds = 1499.13 MB/sec |
33 |
> > Timing buffered disk reads: 174 MB in 3.01 seconds = 57.79 MB/sec |
34 |
> > |
35 |
> > and this are the results __with__ the CPU roaster: |
36 |
> > |
37 |
> > solfire:/home/mccramer>sudo hdparm -Tt /dev/sda |
38 |
> > solfire:/home/mccramer>sudo hdparm -Tt /dev/sda |
39 |
> > /dev/sda: |
40 |
> > Timing cached reads: 2160 MB in 2.10 seconds = 1030.12 MB/sec |
41 |
> > Timing buffered disk reads: 174 MB in 3.03 seconds = 57.41 MB/sec |
42 |
> > |
43 |
> > |
44 |
> > The chached reads dropped by ~469MByte/s. The buffered reads are |
45 |
> > nearly the same. |
46 |
> |
47 |
> The buffered reads are all we care about. They are the actual reads |
48 |
> from the disk to RAM. The cached reads is just a repeated read of the |
49 |
> same sector of the disk, so today is really just a test of your memory |
50 |
> bandwidth. Since we are loading memory and the CPU pretty heavily for |
51 |
> this test, and significant drop is to be expected. |
52 |
|
53 |
Ok...sounds good -- in the sense of: It seems, that I have no |
54 |
hardware problem anywhere in my Linux box... |
55 |
|
56 |
> So, it is not CPU utilization that is hurting your performance. |
57 |
|
58 |
( :) imaging the above sentence *WITHOUT* the current context :) ) |
59 |
|
60 |
This is the eigth wonder of the world...the first time when CPU load |
61 |
does *not* hurt system performance! Oh yeah! I will send all my |
62 |
render tasks to the...floppy controller, hahahahahahaa :)))))) |
63 |
|
64 |
(sorry could not resist...I am a little daft this morning as it seems :O)) |
65 |
|
66 |
> You mentioned problems compiling. The most likely case I can think of |
67 |
> is that you do not have enough memory, and are inducing the system to |
68 |
> swap. |
69 |
|
70 |
Hmmm...1GByte Dualchannel-RAM should be enough for compiling Blender |
71 |
(for example). |
72 |
|
73 |
> Indeed when compiling most programs, you should see very little |
74 |
> if any disk activity. |
75 |
|
76 |
My SATA disk (Seagate ST3200827AS) is heavily shakeing its head when |
77 |
compiling... |
78 |
|
79 |
> This is particularly suspect if you have |
80 |
> something like MAKEOPTS=-j4. |
81 |
|
82 |
Yes, normally I use "make -j 4" for useing both cores. |
83 |
|
84 |
May be I foolishly forget something to switch on or off in my BIOS |
85 |
while migrating from PATA to SATA ? The only PATAs in my system is a |
86 |
Plextor CD reader/burner and a LG DVD reader/burner on IDE1 |
87 |
(scnd. channel). |
88 |
|
89 |
My mobo is a ASUS AV8 with AMI BIOS (upgraded to the "newest" |
90 |
version I could find on the net). |
91 |
|
92 |
> Regards, |
93 |
> -Richard |
94 |
> -- |
95 |
> gentoo-user@g.o mailing list |
96 |
> |
97 |
-- |
98 |
gentoo-user@g.o mailing list |