Re: Analogue of Windows I/O completion ports on Linux?
From: David Schwartz (davids_at_webmaster.com)
Date: 09/24/05
- Next message: Tauno Voipio: "Re: defrag on linux system"
- Previous message: Kenneth: "Re: makefile question / problem"
- In reply to: Erik de Castro Lopo: "Re: Analogue of Windows I/O completion ports on Linux?"
- Next in thread: Peter T. Breuer: "Re: Analogue of Windows I/O completion ports on Linux?"
- Reply: Peter T. Breuer: "Re: Analogue of Windows I/O completion ports on Linux?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sat, 24 Sep 2005 12:26:39 -0700
"Erik de Castro Lopo" <nospam@mega-nerd.com> wrote in message
news:4334EE0C.2B60937A@mega-nerd.com...
> If you want performance you should forget about what windows does and find
> out how to best do it on Linux.
True.
> On Linux I think you will probably maximize
> performance by using blocking I/O with select() in a single thread.
Oh god no. All that has to happen is one client does something unusual
that requires you to fault in some code that hasn't run yet or has been
oppurtunistically paged out. For as long as the disk takes to service the
fault, you will be unable to do anything at all. And you may never catch
back up.
Maximum performance on Linux comes from clever multithreading and use of
epoll. You probably don't need the absolute maximum though.
DS
- Next message: Tauno Voipio: "Re: defrag on linux system"
- Previous message: Kenneth: "Re: makefile question / problem"
- In reply to: Erik de Castro Lopo: "Re: Analogue of Windows I/O completion ports on Linux?"
- Next in thread: Peter T. Breuer: "Re: Analogue of Windows I/O completion ports on Linux?"
- Reply: Peter T. Breuer: "Re: Analogue of Windows I/O completion ports on Linux?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|