Mailing List Archive
tlug.jp Mailing List tlug archive tlug Mailing List Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]Re: [tlug] Re: /proc/uptime
- Date: Fri, 14 Nov 2008 18:45:10 +0900
- From: "Stephen J. Turnbull" <stephen@example.com>
- Subject: Re: [tlug] Re: /proc/uptime
- References: <78d7dd350811111815y11b471ek31c44c4443d2dca5@mail.gmail.com> <87bpwlpqnm.fsf-genuine-vii@john.fremlin.org> <78d7dd350811112048y1bcb2717j19af91dcda1ab257@mail.gmail.com> <877i79pkwz.fsf-genuine-vii@john.fremlin.org> <78d7dd350811112354t3d1555c7s2164a3045373eaf3@mail.gmail.com> <87vdutnk32.fsf-genuine-vii@john.fremlin.org> <78d7dd350811120600i66663e1dka71952f6fa789418@mail.gmail.com> <87vdurow8y.fsf@xemacs.org> <78d7dd350811131931j2f423f28id7b01f3439c71a1@mail.gmail.com>
Nguyen Vu Hung writes: > On Fri, Nov 14, 2008 at 11:51 AM, Stephen J. Turnbull <stephen@example.com>wrote: > > Nguyen Vu Hung writes: > > > > > > No. It indicates the amount of time any one of the cores has > > > > nothing to do. > > > > > Is this a good design? I think no. I will replace OR with AND. > > > Not every recently applications are SMP aware, and some never is. > > > > So? That doesn't change the fact that some of the processors are > > doing almost no useful work! > > I did mean "It indicates the amount of time any one of the cores > has nothing to do" does not make a good sense and it should be "It > indicates the amount of time *all the* cores has nothing to do". Well, yes, I know that's what you meant. It still doesn't make your definition useful. Your definition is biased upward. > > And multi-thread programming (like explicit heap management) is a > > waste of programmer skull-sweat in most server applications > > I take graphicsmagick as an example. Urk. Examples are irrelevant to "most of the time". And look, if you need the CPU, of course you need the CPU. Nobody has contested that. But guess what? The conventional statistic checked when you are running graphicsmagick will show that you need CPU! The point is to find a number where when you need CPU it shows a high load, and when you *don't* need CPU it shows you a *low* load. > > So, if your number says some of your processors are idle 99% of the > > time, and that bothers you, > I dual boot the dual Xeon box. Well, OK, if you buy Intel, you obviously don't care at all how much money you spend on CPU!<wink> > On Windows, I run Adobe PS3, which is using Adobe Camera Raw) to > convert hundreds of Nikon NEF files to JPEG. I really benefit from > SMP in this case. Sure. And that's the point of the precondition "it bothers you". You obviously like having your Windows do its thing really quickly so you can get back to using a real OS as quickly as possible. It evidently *doesn't* bother you to have an 0.01 load level 99% of the time if for the 1% (or whatever) of the time you've got a bunch of pix to process the box screams. In fact, in the case of this kind of image processing it doesn't make sense to check /proc/uptime at all, since what you are most interested in is peak-load performance. But take those guys who are asking about MPUs. They're obviously interested in applications where spending too much money on CPUs is bad, and I suppose there's a good chance they want to just buy enough CPU to ensure that the data gets processed sooner or later (in CPU cycles; in clock time probably a few seconds is the limit, although I can easily imagine applications where it would be OK as long as you catch up every day). Then they might hope to push that ratio as high as 25%.
- References:
- [tlug] /proc/uptime
- From: Nguyen Vu Hung
- [tlug] Re: /proc/uptime
- From: John Fremlin
- Re: [tlug] Re: /proc/uptime
- From: Nguyen Vu Hung
- [tlug] Re: /proc/uptime
- From: John Fremlin
- Re: [tlug] Re: /proc/uptime
- From: Nguyen Vu Hung
- [tlug] Re: /proc/uptime
- From: John Fremlin
- Re: [tlug] Re: /proc/uptime
- From: Nguyen Vu Hung
- Re: [tlug] Re: /proc/uptime
- From: Stephen J. Turnbull
- Re: [tlug] Re: /proc/uptime
- From: Nguyen Vu Hung
Home | Main Index | Thread Index
- Prev by Date: Re: [tlug] Re: /proc/uptime
- Next by Date: Re: [tlug] ARM: from dirt to Linux
- Previous by thread: Re: [tlug] Re: /proc/uptime
- Next by thread: [tlug] Microcontrollers + linux
- Index(es):
Home Page Mailing List Linux and Japan TLUG Members Links