Re: Analogue of Windows I/O completion ports on Linux?
Date: Sat, 24 Sep 2005 12:26:39 -0700
"Erik de Castro Lopo" <email@example.com> wrote in message
> If you want performance you should forget about what windows does and find
> out how to best do it on Linux.
> 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
Maximum performance on Linux comes from clever multithreading and use of
epoll. You probably don't need the absolute maximum though.
- Re: Analogue of Windows I/O completion ports on Linux?
... > out how to best do it on Linux. ... > performance by using blocking I/O with selectin a single thread. ... fault, you will be unable to do anything at all. ...
- Segmentation fault in network-functions
... fault on trying to enable the interface after it has ... to Linux. ... I was told that their government contract requires ...
- Re: Segmentation Fault
... >>You don't want to imply that it is the operating systems fault, ... > call it problematic for portability when Linux actively tries to prevent ... That's the responsibility of the compiler or static analysis tools. ... Instead of blaming the OS you better blame the C programming language ...
- Re: Very OT: any linux users here?
... He is all fault who has no fault at all. ... Then, Charlie, you'll love Linux. ... Microsoft has had the appliance mindset since 95, ... computers and computing from Windows, and will eventually reach a point ...
- RE: What modems work with Ubuntu?
... Or do you insist that Linux should run on a Walmart toaster too? ... Why could they be at fault?! ... It is -their- hardware. ... free o/s on hardware that does not support the free o/s. ...