Re: sendfile(2) SF_NOPUSH flag proposal

From: Peter Jeremy (peterjeremy_at_optushome.com.au)
Date: 05/27/03

  • Next message: Igor Sysoev: "Re: sendfile(2) SF_NOPUSH flag proposal" 
    Date: Tue, 27 May 2003 21:41:10 +1000
To: Igor Sysoev <is@rambler-co.ru>

    On Tue, May 27, 2003 at 03:25:31PM +0400, Igor Sysoev wrote:
    >On Tue, 27 May 2003, Peter Jeremy wrote:
    >
    >> 2) The new feature provides significant performance benefit. In this
    >> case, I believe the overhead of calling setsockopt(2) is negligible
    >> so the performance gain would be negligible.
    >
    >I think the calling setsockopt(TCP_NOPUSH, 1) syscall has huge overhead
    >as compared to several C operators inside sendfile(2).

    Agreed. But needing another one or two relatively cheap system calls
    is negligible compared to the total overhead of accept(), several
    select()/poll()/kqueue() calls, several sendfile() calls and a
    close(). If you can produce some figures demonstrating that two
    setsockopt() syscalls makes a noticable difference to the total cost
    of sending a 48K file then your change might be worth considering.

    >My point is not "the cheap change" but "the cheap overhead".

    Please quantify the improvement amortised over a complete transaction
    or connection.

    Peter
    _______________________________________________
    freebsd-arch@freebsd.org mailing list
    http://lists.freebsd.org/mailman/listinfo/freebsd-arch
    To unsubscribe, send any mail to "freebsd-arch-unsubscribe@freebsd.org"

  • Next message: Igor Sysoev: "Re: sendfile(2) SF_NOPUSH flag proposal"

    Relevant Pages

    • Managed vs Unmanaged Bare Bones Performance Test
      ... just the basic stuff inside the CLR i.e. function calling ... Here are the results (for release mode running from console, ... //record the overhead for calling the performance counter API ...
      (microsoft.public.dotnet.framework.clr)
    • Managed vs Unmanaged Bare Bones Performance Test
      ... just the basic stuff inside the CLR i.e. function calling ... Here are the results (for release mode running from console, ... //record the overhead for calling the performance counter API ...
      (microsoft.public.dotnet.framework)
    • Re: Local variables controversial?
      ... >>Forth is designed so that calling a word has a very low overhead. ... I guess you're assuming that the overhead of inlining a subroutine is ... looping structures are an ANS thing that you don't have in F83. ...
      (comp.lang.forth)
    • Re: Postback from javascript
      ... overhead of calling the iterator on the 'forms' object (although its not ... >> hidden text fields are used to notify the processing page what control ... >> accomplish what you are trying to accomplish with minimal effort. ...
      (microsoft.public.dotnet.framework.aspnet)
    • Re: [RFC] CryptoAPI & Compression
      ... > I think the overhead could be higher. ... then for 1M of incompressible input the total overhead is ... output buffer in order to let zlib finish its stream. ... send the line "unsubscribe linux-kernel" in ...
      (Linux-Kernel)