Re: TOP interpretation

One quick question Brad.
Do you know of a way of displaying the CPU time separated by CPUs like
HP-UX top?
I saw that <I> option that talks about Irix/Solaris mode but I'm not
sure still how to use it. I was hoping to be able to see the usage per
each core, when using a multicore CPU like the new AMD CPUs.

TIA,

Dumitru

On Fri, Jun 26, 2009 at 5:47 PM, Brad Barnett<[email protected]> wrote:
>
>
> Ah, you've found a RTFM escape clause -- or at least a hall pass. :P
>
> If you're really stuck, however, there are tons and tons of webpages
> dedicated to manpages.  There is also Google to help you, too.
>
> Funny thing, though:
>
> https://bugzilla.redhat.com/show_bug.cgi?id=296471
>
> It seems the section you are looking for, is hard to find in several
> distros.  One of my Debian boxes is missing section 2c, and it looks like
> several Redhat (and therefore CentOS) is too.
>
> Check this page, with section 2c in it:
>
> http://www.annodex.net/cgi-bin/man/man2html?1+top
>
> On Fri, 26 Jun 2009 13:29:49 -0400
> OddSox <[email protected]> wrote:
>
>> I don't know whether I should top or bottom post at this point.
>>
>> I had sent an earlier post where I tried "man top" and got crap that
>> I couldn't read because of encoding issues, so I posted the later
>> message. Perhaps the earlier one didn't make it to the list?
>>
>> > How do I fix this crap?
>> >
>> > -bash-3.2$ man top - results in gibberish like...
>> >
>> >      13        The traditional switches ?<80><99>-?<80><99> and
>> > whitespace are optional.
>> >
>> > This particular box is a CentOS 5.x server, although I've had the
>> > same thing happen on many different systems for years and I'm
>> > getting tired of it. Asking for "man" on just about anything
>> > results in unreadable stuff related to unicode or something. I only
>> > use terminal so there must be some setting I can hit for this?
>> >
>> > Is it because I'm using bash?
>>
>>
>>
>> On Jun 26, 2009, at 12:51 PM, Brad Barnett wrote:
>>
>> >
>> > man 'top' should help you, Oddsox.
>> >
>> > It will clarify much below, and correct the "wait" statement below.
>> >
>> > That is:
>> >
>> >         wa  --  iowait
>> >           Amount of time the CPU has been waiting for I/O to complete.
>> >
>> > See more info in the docs, or put another way ... RTFM :P
>> >
>> > On Fri, 26 Jun 2009 12:33:10 -0400
>> > Damian Gerow <[email protected]> wrote:
>> >
>> >> OddSox wrote:
>> >> : Further to my last message, I'm trying to figure out the
>> >> significance : of the various CPU numbers in TOP on my CentOS 5.x
>> >> server... :
>> >> : > Cpu(s):  0.3%us,  0.0%sy,  0.0%ni, 99.7%id,  0.0%wa,  0.0%hi,
>> >> 0.0%si,  0.0%st :
>> >> :
>> >> : What does 99.7%id mean?
>> >>
>> >> %us = user
>> >> %sy = system
>> >> %ni = nice
>> >> %id = idle
>> >> %wa = wait
>> >> %hi = ?
>> >> %si = system interrupts?
>> >> %st = ?
>> >>
>> >> I've never seen those last three, but it's been a few years since I
>> >> bothered with 'top' in Linux.  System performance is a notoriously
>> >> difficult thing to measure and understand...
>> >>
>> >> user = amount of time the system is running userland code
>> >> system = amount of time the system is running kernel code
>> >> nice = I've never really understood this one
>> >> idle = amount of time the system is doing nothing
>> >> wait = amount of time the system is waiting for available CPU cycles
>> >>
>> >> Or something like that.
>> >> --
>> >> OCLUG general discussion list
>> >> [email protected]
>> >> http://oclug.on.ca/mailman/listinfo/oclug
>> > --
>> > OCLUG general discussion list
>> > [email protected]
>> > http://oclug.on.ca/mailman/listinfo/oclug
>>
>> --
>> OCLUG general discussion list
>> [email protected]
>> http://oclug.on.ca/mailman/listinfo/oclug
> --
> OCLUG general discussion list
> [email protected]
> http://oclug.on.ca/mailman/listinfo/oclug
>



-- 
=====================

Dumitru Ciobanu
[email protected]
-- 
OCLUG general discussion list
[email protected]
http://oclug.on.ca/mailman/listinfo/oclug