comp.os.linux.misc
http://groups.google.com/group/comp.os.linux.misc?hl=en
comp.os.linux.misc@googlegroups.com
Today's topics:
* Xfce's screen saver - 7 messages, 5 authors
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/421e7ed6a3fc2d8c?hl=en
* opacity of Ubuntu - 4 messages, 3 authors
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/d9c960920b115b36?hl=en
* NYC LOCAL: Tuesday 11 November 2008 Lisp NYC Eating and Drinking and Talking
- 1 messages, 1 author
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/dd660930f535aefc?hl=en
* linux energy usage - 5 messages, 5 authors
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/7eae86e585d31b91?hl=en
* Intel Quad Core performance on CentOS 5.2 is slower - 6 messages, 4 authors
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/441111eb265f9167?hl=en
* losetup -d LOOP_CLR_FD: Device or resource busy - 1 messages, 1 author
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/e94540c418eb3e3e?hl=en
* Need a shell script - 1 messages, 1 author
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/af92867b3096c7a8?hl=en
==============================================================================
TOPIC: Xfce's screen saver
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/421e7ed6a3fc2d8c?hl=en
==============================================================================
== 1 of 7 ==
Date: Mon, Nov 10 2008 8:46 pm
From: Sidney Lambe
["Followup-To:" header set to comp.os.linux.misc.]
Hactar <ebenZEROONE@verizon.net> wrote:
> I have Ubuntu 8.04, with Xfce as the environment in X. I cannot get
> Xfce's screen saver to engage. X's (as controlled with xset) works fine,
> but it's boring. DPMS works, except that if I change xfce's control for
> it, it gets disabled (all times in xset go to 0). Anyone know a solution
> to either of these problems?
So you are worried about what's on the screen when you
aren't looking at it?
I just turn my monitor off when I'm not using it.
Can't see any point in wasting energy on something
that I'm not using.
No one has ever complained about the blank screen.
If they did, I'd think they had a loose screw or two.
Sid
== 2 of 7 ==
Date: Tues, Nov 11 2008 6:53 am
From: ebenZEROONE@verizon.net (Hactar)
In article <bafnu5x8i1.ln2@amma.net>,
Sidney Lambe <nospam@nospam.invalid> wrote:
> ["Followup-To:" header set to comp.os.linux.misc.]
> Hactar <ebenZEROONE@verizon.net> wrote:
> > I have Ubuntu 8.04, with Xfce as the environment in X. I cannot get
> > Xfce's screen saver to engage. X's (as controlled with xset) works fine,
> > but it's boring. DPMS works, except that if I change xfce's control for
> > it, it gets disabled (all times in xset go to 0). Anyone know a solution
> > to either of these problems?
>
> So you are worried about what's on the screen when you
> aren't looking at it?
Just because I'm not sitting at the computer doesn't mean I don't
occasionally glance at it.
> I just turn my monitor off when I'm not using it.
> Can't see any point in wasting energy on something
> that I'm not using.
I have DPMS kick in at a half hour; same deal.
> No one has ever complained about the blank screen.
>
> If they did, I'd think they had a loose screw or two.
All black = good for CRTs, but worst-case (in terms of power
consumption) for LCDs.
--
-eben QebWenE01R@vTerYizUonI.nOetP royalty.mine.nu:81
PISCES: Try to avoid any Virgos or Leos with the Ebola
virus. You are the Lord of the Dance, no matter what those
idiots at work say. -- Weird Al, _Your Horoscope for Today_
== 3 of 7 ==
Date: Tues, Nov 11 2008 7:59 am
From: Michael Black
On Tue, 11 Nov 2008, Hactar wrote:
> In article <bafnu5x8i1.ln2@amma.net>,
> Sidney Lambe <nospam@nospam.invalid> wrote:
>> ["Followup-To:" header set to comp.os.linux.misc.]
>> Hactar <ebenZEROONE@verizon.net> wrote:
>>> I have Ubuntu 8.04, with Xfce as the environment in X. I cannot get
>>> Xfce's screen saver to engage. X's (as controlled with xset) works fine,
>>> but it's boring. DPMS works, except that if I change xfce's control for
>>> it, it gets disabled (all times in xset go to 0). Anyone know a solution
>>> to either of these problems?
>>
>> So you are worried about what's on the screen when you
>> aren't looking at it?
>
> Just because I'm not sitting at the computer doesn't mean I don't
> occasionally glance at it.
>
>> I just turn my monitor off when I'm not using it.
>> Can't see any point in wasting energy on something
>> that I'm not using.
>
> I have DPMS kick in at a half hour; same deal.
>
>> No one has ever complained about the blank screen.
>>
>> If they did, I'd think they had a loose screw or two.
>
> All black = good for CRTs, but worst-case (in terms of power
> consumption) for LCDs.
>
Of course, screensavers were about blanking the screen so a fixed pattern
wasn't constantly on the screen; done long enough with the same pattern
and it can burn into the screen. I used to have an old point of sale
terminal that had some burn in over in the corner, and certainly the rise
of desktops on computers means a very similar pattern is there constantly.
But somewhere along the way, screensavers became "art". Yes, they do
ensure that the same pattern is not seen by the screen all the time, but
they sure don't do anything about saving power. IN order to display all
those fancy flying toasters and whatever, the monitor has to be in
constant use.
I sure don't bother with a screensaver. The default on the distribution I
use blanks the screen after a certain amount of time, and then puts the
monitor in standbye after a longer certain length of time.
Michael
== 4 of 7 ==
Date: Tues, Nov 11 2008 9:06 am
From: ebenZEROONE@verizon.net (Hactar)
In article <Br6Sk.1813$A73.616@edtnps82>,
johnny bobby bee <useraddshine-no_spam_eh@yahoo.ca> wrote:
> Hactar wrote:
> > I have Ubuntu 8.04, with Xfce as the environment in X. I cannot get
> > Xfce's screen saver to engage. X's (as controlled with xset) works fine,
> > but it's boring. DPMS works, except that if I change xfce's control for
> > it, it gets disabled (all times in xset go to 0). Anyone know a solution
> > to either of these problems?
>
> Can't help with the Xfce screensaver.
>
> But, how is this boring:
> xset s 1 1
> Pretty sweet if you ask me. ;)
>
> Have you installed xscreensaver and played with the
> 'xscreensaver-command' or 'xscreensaver-demo' commands?
xscreensaver's nice, I had it in my former installation. But I hate
duplicated functionality. Is there a way of uninstalling xfce's screen
saver, or do they use the same guts? Some of the module names are the
same, but I don't think that necessarily means the _files_ are the same.
--
-eben QebWenE01R@vTerYizUonI.nOetP royalty.mine.nu:81
Two atoms are discussing whether or not to go into a bar. The first atom
says, "I don't know about this place. Last time I went in there, I lost
an electron." The second atom says, "Don't worry, I'll keep my ion you."
== 5 of 7 ==
Date: Tues, Nov 11 2008 9:20 am
From: ebenZEROONE@verizon.net (Hactar)
In article <Pine.LNX.4.64.0811111055220.9150@darkstar.example.net>,
Michael Black <et472@ncf.ca> wrote:
> On Tue, 11 Nov 2008, Hactar wrote:
>
> > In article <bafnu5x8i1.ln2@amma.net>,
> > Sidney Lambe <nospam@nospam.invalid> wrote:
> >
> >> I just turn my monitor off when I'm not using it.
> >> Can't see any point in wasting energy on something
> >> that I'm not using.
> >
> > I have DPMS kick in at a half hour; same deal.
> >
> >> No one has ever complained about the blank screen.
> >>
> >> If they did, I'd think they had a loose screw or two.
> >
> > All black = good for CRTs, but worst-case (in terms of power
> > consumption) for LCDs.
> >
> Of course, screensavers were about blanking the screen so a fixed pattern
> wasn't constantly on the screen; done long enough with the same pattern
> and it can burn into the screen. I used to have an old point of sale
> terminal that had some burn in over in the corner, and certainly the rise
> of desktops on computers means a very similar pattern is there constantly.
>
> But somewhere along the way, screensavers became "art".
Maybe when "Pyro" came out for the Mac, sometime in the mid-late 80s?
> Yes, they do
> ensure that the same pattern is not seen by the screen all the time, but
> they sure don't do anything about saving power.
Well, a CRT displaying mostly black does use a fair amount less power
than one displaying a lighter workspace. To be sure, it's still a lot
more than "off".
> I sure don't bother with a screensaver. The default on the distribution I
> use blanks the screen after a certain amount of time, and then puts the
> monitor in standbye after a longer certain length of time.
On/off too often isn't very good for LCDs (the backlight is a
fluorescent tube or two). I'm mostly interested that it be not my
current screen, and interesting to look at. Pretty pictures are like
wall-art that moves, in my book.
--
-eben QebWenE01R@vTerYizUonI.nOetP royalty.mine.nu:81
A neutrino enters a bar. The bartender asks "Can I get you anything?"
The neutrino replies "Nope, just passing through." -- rdu_voyager on Fark
== 6 of 7 ==
Date: Tues, Nov 11 2008 10:38 am
From: Bill Marcum
["Followup-To:" header set to comp.os.linux.misc.]
On 2008-11-11, Hactar <ebenZEROONE@verizon.net> wrote:
>>
>> Have you installed xscreensaver and played with the
>> 'xscreensaver-command' or 'xscreensaver-demo' commands?
>
> xscreensaver's nice, I had it in my former installation. But I hate
> duplicated functionality. Is there a way of uninstalling xfce's screen
> saver, or do they use the same guts? Some of the module names are the
> same, but I don't think that necessarily means the _files_ are the same.
>
xfce installs gnome-screensaver. If you uninstall that, it might say
that xfce-desktop will be removed, but that is just a metapackage.
--
They can't stop us... we're on a mission from God!
-- The Blues Brothers
== 7 of 7 ==
Date: Tues, Nov 11 2008 10:34 am
From: Balwinder S Dheeman
On 11/11/2008 08:23 PM, Hactar wrote:
> In article <bafnu5x8i1.ln2@amma.net>,
> Sidney Lambe <nospam@nospam.invalid> wrote:
>> ["Followup-To:" header set to comp.os.linux.misc.]
>> Hactar <ebenZEROONE@verizon.net> wrote:
>>> I have Ubuntu 8.04, with Xfce as the environment in X. I cannot get
>>> Xfce's screen saver to engage. X's (as controlled with xset) works fine,
>>> but it's boring. DPMS works, except that if I change xfce's control for
>>> it, it gets disabled (all times in xset go to 0). Anyone know a solution
>>> to either of these problems?
>> So you are worried about what's on the screen when you
>> aren't looking at it?
>
> Just because I'm not sitting at the computer doesn't mean I don't
> occasionally glance at it.
>
>> I just turn my monitor off when I'm not using it.
>> Can't see any point in wasting energy on something
>> that I'm not using.
>
> I have DPMS kick in at a half hour; same deal.
>
>> No one has ever complained about the blank screen.
>>
>> If they did, I'd think they had a loose screw or two.
>
> All black = good for CRTs, but worst-case (in terms of power
> consumption) for LCDs.
I think, the manual page for xorg.conf has enough of information, here
are requisite excerpts from relevant sections:
Section "ServerFlags"
...
Option "BlankTime" "5" # in minutes
Option "StandbyTime" "5" # vsync, nearly instant recovery
Option "SuspendTime" "5" # hsync, typically 3s recovery
Option "OffTime" "5" # poff, typically 10s recovery
EndSection
Section "Monitor"
...
Option "DPMS" "true"
EndSection
All you need to do is add the above said options and, or change the same
per your needs; this generally is applicable to X server -- display or
video drivers and not specific for XFCE, KDE and GNOME etcetera.
Please note, that the LCD monitors recover quite quickly compared to the
CRT monitors.
IMHO, only blanking a display is the best screen-saver ever made.
Hope that helps,
--
Dr Balwinder S "bsd" Dheeman Registered Linux User: #229709
Anu'z Linux@HOME (Unix Shoppe) Machines: #168573, 170593, 259192
Chandigarh, UT, 160062, India Gentoo, Fedora, Debian/FreeBSD/XP
Home: http://cto.homelinux.net/~bsd/ Visit: http://counter.li.org/
==============================================================================
TOPIC: opacity of Ubuntu
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/d9c960920b115b36?hl=en
==============================================================================
== 1 of 4 ==
Date: Tues, Nov 11 2008 1:49 am
From: rabiticide
I am having the same frustrations with Ubuntu that I used to feel
towards Windows. Is it just me, or did they make Ubuntu hard to
Monkey with?
My specific problem is this: My "add/remove software" option just
disappeared and so did my synaptic package manager. This has to be
something easy. What is the folder which contain links from the GUI
"Applications" and "system" drop ups?
rK
== 2 of 4 ==
Date: Tues, Nov 11 2008 7:07 am
From: Bill Marcum
On 2008-11-11, rabiticide <rabiticide@gmail.com> wrote:
> I am having the same frustrations with Ubuntu that I used to feel
> towards Windows. Is it just me, or did they make Ubuntu hard to
> Monkey with?
>
> My specific problem is this: My "add/remove software" option just
> disappeared and so did my synaptic package manager. This has to be
> something easy. What is the folder which contain links from the GUI
> "Applications" and "system" drop ups?
>
> rK
You can still use the command line to run synaptic, unless it's been
uninstalled, in that case you can get it back with apt-get. I'm not sure
what the real name of the "add/remove software" app is, but it shouldn't
be too hard to find out.
--
They can't stop us... we're on a mission from God!
-- The Blues Brothers
== 3 of 4 ==
Date: Tues, Nov 11 2008 8:14 am
From: propman
rabiticide wrote:
> I am having the same frustrations with Ubuntu that I used to feel
> towards Windows. Is it just me, or did they make Ubuntu hard to
> Monkey with?
>
> My specific problem is this: My "add/remove software" option just
> disappeared and so did my synaptic package manager. This has to be
> something easy. What is the folder which contain links from the GUI
> "Applications" and "system" drop ups?
>
> rK
Right click on the bar --> Edit Menus
== 4 of 4 ==
Date: Tues, Nov 11 2008 11:16 am
From: rabiticide
weird behavior
I did the right click -> edit menu like you said and I saw the
options I wanted. I checked "add/remove" and two seconds later my
check mark disappeared!? I did the same thing with synaptic package
manager - I checked the box, count to two, and my check disappears. I
tried "apt-get install synaptic" as root and it tells me it's already
installed, so that's not the problem...
rK
==============================================================================
TOPIC: NYC LOCAL: Tuesday 11 November 2008 Lisp NYC Eating and Drinking and
Talking
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/dd660930f535aefc?hl=en
==============================================================================
== 1 of 1 ==
Date: Tues, Nov 11 2008 2:00 am
From: secretary@lxny.org
<blockquote
what="official Lisp NYC announcement"
edits="">
Date: Mon, 10 Nov 2008 16:36:41 -0500
From: "H Goodman" <hg574425@gmail.com>
To: lisp@lispnyc.org
Join us for our next social meeting on Tuesday, November 11th from
7:00 to 9:00 at Westside Brewery 340 Amsterdam Ave (From the 1,2,3,9:
between 76th-77th on Amsterdam)
Cheers,
- Heow
_______________________________________________
Lisp mailing list
Lisp@lispnyc.org
http://www.lispnyc.org:8080/mailman/listinfo/lisp
</blockquote>
Distributed poC TINC:
Jay Sulzberger <secretary@lxny.org>
Corresponding Secretary LXNY
LXNY is New York's Free Computing Organization.
http://www.lxny.org
==============================================================================
TOPIC: linux energy usage
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/7eae86e585d31b91?hl=en
==============================================================================
== 1 of 5 ==
Date: Tues, Nov 11 2008 2:14 am
From: russ
On 11 Nov, 01:13, Jean-David Beyer <jeandav...@verizon.net> wrote:
> russ wrote:
> > Hi
>
> > is there a [s]imple way in software to monitor the energy usage of a
> > linux machine? can one monitor fan voltage, cpu/disk voltage and attempt
> > to try and infer overall computer energy use?
>
> The way I do it, since I do not really care much, is to look at the
> indicator lights on my UPS. If you go here, and enlarge the picture, you can
> see the lights on the left side of the unit (as you face it).
> If they are all off, you are using less than 17% of the power it can supply.
> If the bottom one is on, you are using between 17% and 33%
> If the bottom two are on, you are using between 33% and 50%
> If the bottom three are on, you are using between 50% and 67%
> If the bottom four are on, you are using between 67% and 85%
> If they are all on, you are using over 85%. Since my unit is rated at 2200
> VA, I can easily figure it out.
>
> The lights on the other side are charge in the battery, but they are easier
> to remember: 20%, 40%, 60%, 80% and 100%.
>
> > thanks in advance
>
> You could also get a power meter, as has been suggested by another poster.
> That one would not work for me because my UPS uses a 20Amp plug, and that
> meter only accepts a 15Amp plug.
thanks all for the above. it largely confirms what i thought, that no
you can't do it in software. i was hoping that i could use cpu usage,
fan, cpu temperature or other proxies to estimate energy usage. the
plug in monitor is fine for my computer but I want to monitor several
hundred to see the energy they are putting into a building (i am an
energy rather than computer geek). is it possible to find the rail
voltage in software? is the fan a good proxy for energy use and is it
controlled by the OS?
russ
== 2 of 5 ==
Date: Tues, Nov 11 2008 3:43 am
From: TheLostPacket
On Nov 11, 10:14 am, russ <rlaybe...@hotmail.com> wrote:
> On 11 Nov, 01:13, Jean-David Beyer <jeandav...@verizon.net> wrote:
>
>
>
> > russ wrote:
> > > Hi
>
> > > is there a [s]imple way in software to monitor the energy usage of a
> > > linux machine? can one monitor fan voltage, cpu/disk voltage and attempt
> > > to try and infer overall computer energy use?
>
> > The way I do it, since I do not really care much, is to look at the
> > indicator lights on my UPS. If you go here, and enlarge the picture, you can
> > see the lights on the left side of the unit (as you face it).
> > If they are all off, you are using less than 17% of the power it can supply.
> > If the bottom one is on, you are using between 17% and 33%
> > If the bottom two are on, you are using between 33% and 50%
> > If the bottom three are on, you are using between 50% and 67%
> > If the bottom four are on, you are using between 67% and 85%
> > If they are all on, you are using over 85%. Since my unit is rated at 2200
> > VA, I can easily figure it out.
>
> > The lights on the other side are charge in the battery, but they are easier
> > to remember: 20%, 40%, 60%, 80% and 100%.
>
> > > thanks in advance
>
> > You could also get a power meter, as has been suggested by another poster.
> > That one would not work for me because my UPS uses a 20Amp plug, and that
> > meter only accepts a 15Amp plug.
>
> thanks all for the above. it largely confirms what i thought, that no
> you can't do it in software. i was hoping that i could use cpu usage,
> fan, cpu temperature or other proxies to estimate energy usage. the
> plug in monitor is fine for my computer but I want to monitor several
> hundred to see the energy they are putting into a building (i am an
> energy rather than computer geek). is it possible to find the rail
> voltage in software? is the fan a good proxy for energy use and is it
> controlled by the OS?
>
> russ
most BIOSes have rail voltage/fan monitors, I'm pretty sure there
should be some way (eg on a Windows box, with MBM) you can monitor the
output of that in Linux. Don't ask me what though, 'cos I don't know.
Hardware wise, I think someone already pointed a link to an inline
power meter which does realtime monitoring of input power, voltage,
and the total KWh used while the thing's had power going through it.
Very nifty device, and yes it does give surprising results (like, a
VIA C3 Nehemiah 1.0GHz on an Epia M board, with an 80GB notebook
drive, no fans, 22" 65W widescreen and Sky Digital box giving change
out of 100W!)
TLP
== 3 of 5 ==
Date: Tues, Nov 11 2008 4:17 am
From: Maxwell Lol
russ <rlayberry@hotmail.com> writes:
> Hi
>
> is there a aimple way in software to monitor the energy usage of a
> linux machine? can one monitor fan voltage, cpu/disk voltage and
> attempt to try and infer overall computer energy use?
Buy a device called Kill-A-Watt, and plug your computer into it.
== 4 of 5 ==
Date: Tues, Nov 11 2008 6:56 am
From: ebenZEROONE@verizon.net (Hactar)
In article <4070fdfa-5659-4cd1-bdb4-ee86437d71f5@k36g2000pri.googlegroups.com>,
russ <rlayberry@hotmail.com> wrote:
>
> thanks all for the above. it largely confirms what i thought, that no
> you can't do it in software. i was hoping that i could use cpu usage,
> fan, cpu temperature or other proxies to estimate energy usage. the
> plug in monitor is fine for my computer but I want to monitor several
> hundred to see the energy they are putting into a building (i am an
> energy rather than computer geek). is it possible to find the rail
> voltage in software?
I believe you can get apcaccess to ask apcupsd to ask the UPS what the
line voltage is, but I don't know any way of asking the computer
directly.
--
-eben QebWenE01R@vTerYizUonI.nOetP royalty.mine.nu:81
This message was created using recycled electrons.
== 5 of 5 ==
Date: Tues, Nov 11 2008 9:54 am
From: Jean-David Beyer
Hactar wrote:
> In article <4070fdfa-5659-4cd1-bdb4-ee86437d71f5@k36g2000pri.googlegroups.com>,
> russ <rlayberry@hotmail.com> wrote:
>> thanks all for the above. it largely confirms what i thought, that no
>> you can't do it in software. i was hoping that i could use cpu usage,
>> fan, cpu temperature or other proxies to estimate energy usage. the
>> plug in monitor is fine for my computer but I want to monitor several
>> hundred to see the energy they are putting into a building (i am an
>> energy rather than computer geek). is it possible to find the rail
>> voltage in software?
>
> I believe you can get apcaccess to ask apcupsd to ask the UPS what the
> line voltage is, but I don't know any way of asking the computer
> directly.
>
True, but you cannot ask it the current, so you cannot computer the power
(energy).
--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ PGP-Key: 9A2FC99A Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 10:15:01 up 4 days, 19:59, 4 users, load average: 4.33, 4.35, 4.34
==============================================================================
TOPIC: Intel Quad Core performance on CentOS 5.2 is slower
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/441111eb265f9167?hl=en
==============================================================================
== 1 of 6 ==
Date: Tues, Nov 11 2008 4:26 am
From: Maxwell Lol
viz <vizziee@gmail.com> writes:
> However, the system is slow while running any application. When I
> check the processor usage in the system monitor, it shows only one
> core being used most of the time. Load on other cores is 1-3% during
> this time.
If you are only using one application, then multiple cores won't help
too much. Most applicatiyons need to be re-written to use multiple
cores (i.e. Photoshop, etc.)
== 2 of 6 ==
Date: Tues, Nov 11 2008 7:51 am
From: Steve Wampler
viz wrote:
> I installed CentOS 5.2 (kernel 2.6.18-92.1.17.el5) on this machine. I
> expected the system to be faster compared to my another machine (Intel
> P4 CPU 2.53 GHz, 1GB RAM) assuming it makes effective use of 4 cores.
>
> However, the system is slow while running any application. When I
> check the processor usage in the system monitor, it shows only one
> core being used most of the time. Load on other cores is 1-3% during
> this time.
Cat /proc/cpuinfo and count the number of cores seen by the OS. From the
above, it looks as though you've installed a single-core version of
the kernel. Traditionally, SMP (multi-core) versions of the kernel are
tagged with 'smp', as in: "2.6.9-67.0.22.ELsmp".
--
Steve Wampler -- swampler@noao.edu
The gods that smiled on your birth are now laughing out loud.
== 3 of 6 ==
Date: Tues, Nov 11 2008 7:57 am
From: Steve Wampler
Steve Wampler wrote:
> Cat /proc/cpuinfo and count the number of cores seen by the OS. From the
> above, it looks as though you've installed a single-core version of
> the kernel. Traditionally, SMP (multi-core) versions of the kernel are
> tagged with 'smp', as in: "2.6.9-67.0.22.ELsmp".
Nevermind - a reread of your post shows you know it is running SMP...
--
Steve Wampler -- swampler@noao.edu
The gods that smiled on your birth are now laughing out loud.
== 4 of 6 ==
Date: Tues, Nov 11 2008 7:55 am
From: Steve Wampler
Steve Wampler wrote:
> Cat /proc/cpuinfo and count the number of cores seen by the OS. From the
> above, it looks as though you've installed a single-core version of
> the kernel. Traditionally, SMP (multi-core) versions of the kernel are
> tagged with 'smp', as in: "2.6.9-67.0.22.ELsmp".
To follow up a bit more. The command:
-> grep processor /proc/cpuinfo
processor : 0
processor : 1
processor : 2
processor : 3
processor : 4
processor : 5
processor : 6
processor : 7
will show you the cores seen by the kernel. And on CentOS 5, the command
uname -a
will tell you if the kernel is multi-core aware as (e.g):
->uname -a
Linux presto.tuc.noao.edu 2.6.18-92.1.13.el5 #1 SMP Wed Sep 24 19:32:05 EDT 2008 x86_64 x86_64 x86_64 GNU/Linux
--
Steve Wampler -- swampler@noao.edu
The gods that smiled on your birth are now laughing out loud.
== 5 of 6 ==
Date: Tues, Nov 11 2008 9:19 am
From: Robert Heller
At Tue, 11 Nov 2008 08:51:29 -0700 Steve Wampler <swampler@noao.edu> wrote:
>
> viz wrote:
> > I installed CentOS 5.2 (kernel 2.6.18-92.1.17.el5) on this machine. I
> > expected the system to be faster compared to my another machine (Intel
> > P4 CPU 2.53 GHz, 1GB RAM) assuming it makes effective use of 4 cores.
> >
> > However, the system is slow while running any application. When I
> > check the processor usage in the system monitor, it shows only one
> > core being used most of the time. Load on other cores is 1-3% during
> > this time.
>
> Cat /proc/cpuinfo and count the number of cores seen by the OS. From the
> above, it looks as though you've installed a single-core version of
> the kernel. Traditionally, SMP (multi-core) versions of the kernel are
> tagged with 'smp', as in: "2.6.9-67.0.22.ELsmp".
Note: Most (all?) common desktop applications are single-threaded, and
most would NOT benifit from being rewritten to be multi-threaded. A
multi-core / multi-processor system is a waste of effort for a *Desktop*
machine, unless the machine is making use of specialized applications
that really make use of SMP such as programs like Matlab or various other
number crunching and/or similar scientific applications.
The only other point of a desktop system with multiple cores / multiple
processors would be if one was running something like multiple
instances of SETI-at-home or something thing similar. OR if the
desktop system is also being used as a server of some sort (mail, file,
IntraNet, DHCP, proxy, etc.) such as for a small home [office] LAN.
Also: the amount of RAM and the speed of the disk can have a large
effect on percieved computer speed. Often these factors are more
significant then most people realize. Throwing more cores/processors
into a memory starved system (or a system with a slow disk), will NOT
make it faster.
>
--
Robert Heller -- Get the Deepwoods Software FireFox Toolbar!
Deepwoods Software -- Linux Installation and Administration
http://www.deepsoft.com/ -- Web Hosting, with CGI and Database
heller@deepsoft.com -- Contract Programming: C/C++, Tcl/Tk
== 6 of 6 ==
Date: Tues, Nov 11 2008 10:21 am
From: Jean-David Beyer
Robert Heller wrote:
> At Tue, 11 Nov 2008 08:51:29 -0700 Steve Wampler <swampler@noao.edu>
> wrote:
>
>> viz wrote:
>>> I installed CentOS 5.2 (kernel 2.6.18-92.1.17.el5) on this machine.
>>> I expected the system to be faster compared to my another machine
>>> (Intel P4 CPU 2.53 GHz, 1GB RAM) assuming it makes effective use of 4
>>> cores.
>>>
>>> However, the system is slow while running any application. When I
>>> check the processor usage in the system monitor, it shows only one
>>> core being used most of the time. Load on other cores is 1-3% during
>>> this time.
>> Cat /proc/cpuinfo and count the number of cores seen by the OS. From
>> the above, it looks as though you've installed a single-core version of
>> the kernel. Traditionally, SMP (multi-core) versions of the kernel
>> are tagged with 'smp', as in: "2.6.9-67.0.22.ELsmp".
>
> Note: Most (all?) common desktop applications are single-threaded, and
> most would NOT benifit from being rewritten to be multi-threaded. A
> multi-core / multi-processor system is a waste of effort for a *Desktop*
> machine, unless the machine is making use of specialized applications
> that really make use of SMP such as programs like Matlab or various other
> number crunching and/or similar scientific applications.
I do not know if mine is a "desktop" or not; it is on the top of my desk.
One application I run is postgreSQL dbms. While most of its processing uses
a single processor, its logger, its memory cleanup, and such do use the
other processors, so there is a benefit to multiple cores.
When I ran IBM's DB2 dbms, they ran a disk write process for each hard
drive, so that that could go in parallel with the other stuff.
>
> The only other point of a desktop system with multiple cores / multiple
> processors would be if one was running something like multiple instances
> of SETI-at-home or something thing similar. OR if the desktop system is
> also being used as a server of some sort (mail, file, IntraNet, DHCP,
> proxy, etc.) such as for a small home [office] LAN.
Depending on what the machine is doing, you can get some benefit from more
than one core for ordinary desktop usage. Even if the user's process takes
100% of a CPU, the system and any other daemons, can run on another.
I happen to run 4 BOINC processes to use up the waste processor time.
>
> Also: the amount of RAM and the speed of the disk can have a large effect
> on percieved computer speed. Often these factors are more significant
> then most people realize. Throwing more cores/processors into a memory
> starved system (or a system with a slow disk), will NOT make it faster.
>
Absolutely. That is why I put 8 GBytes in this machine. Right now down to
four due to a memory problem making an error every few days: corrected by
ECC, but I don't like it.
--
.~. Jean-David Beyer Registered Linux User 85642.
/V\ PGP-Key: 9A2FC99A Registered Machine 241939.
/( )\ Shrewsbury, New Jersey http://counter.li.org
^^-^^ 13:15:01 up 4 days, 22:59, 4 users, load average: 4.09, 4.29, 4.50
==============================================================================
TOPIC: losetup -d LOOP_CLR_FD: Device or resource busy
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/e94540c418eb3e3e?hl=en
==============================================================================
== 1 of 1 ==
Date: Tues, Nov 11 2008 6:06 am
From: Scooter
Yeah I've just been using the next loop. I guess that will work until
I can reboot.
==============================================================================
TOPIC: Need a shell script
http://groups.google.com/group/comp.os.linux.misc/browse_thread/thread/af92867b3096c7a8?hl=en
==============================================================================
== 1 of 1 ==
Date: Tues, Nov 11 2008 10:52 am
From: Geico Caveman
pk wrote:
> On Monday 10 November 2008 20:49, Geico Caveman wrote:
>
>>> Ok, then:
>>>
>>> sed 's|\\added\[[^]]*\]{\([^}]*\)}|\1|g
>>> s|\\replaced\[[^]]*\]{\([^}]*\)}{[^}]*}|\1|g
>>> s|\\deleted[[^]]*]{[^}]*}||g' file.tex
>>>
>>> This assumes that no } appears inside {..}, and no ] inside [..] (not
>>> even escaped or in any other form). If you need that, then it becomes
>>> much more complicated.
>>
>> Thanks. I will try this out and see if it works. I was wise to avoid
>> figuring this syntax out by myself :)
>
> btw, there is an error in the last one, should be
>
> s|\\deleted\[[^]]*\]{[^}]*}||g' file.tex
Got that. I am now trying to extend this script to do a few additional
things -
- replace all instances of \marginpar{<text1>} by blank.
I added the following code fragment to the script :
s|\\marginpar{.*}||g
This is not working out (some unmatched sections of the text are
disappearing). I have some familiarity with the :1,$ s/text1/text2/gc
functionality of vim, so was just trying to see if that would work.
What am I doing wrong above ?
==============================================================================
You received this message because you are subscribed to the Google Groups "comp.os.linux.misc"
group.
To post to this group, visit http://groups.google.com/group/comp.os.linux.misc?hl=en
To unsubscribe from this group, send email to comp.os.linux.misc+unsubscribe@googlegroups.com
To change the way you get mail from this group, visit:
http://groups.google.com/group/comp.os.linux.misc/subscribe?hl=en
To report abuse, send email explaining the problem to abuse@googlegroups.com
==============================================================================
Google Groups: http://groups.google.com/?hl=en
0 comments:
Post a Comment