Re: POSIX aio vs completion ports

Stephen C. Tweedie (sct@redhat.com)
Wed, 15 Sep 1999 16:37:12 +0100 (BST)


Hi,

On Tue, 14 Sep 1999 14:31:47 -0400 (EDT), Chuck Lever <cel@monkey.org> said:

>> As far as I know, on a queue overflow, SIGIO is raised and you can
>> select() or poll() on your fds. This of course is going to be
>> exceptionally slow on a large number of descriptors, but it
>> shouldn't happen often.

> i think that kind of design is completely unrealistic. you are most
> likely to run out of queue space when the server is overloaded. why then
> would you want to use a recovery mechanism that would just make the
> overload worse?

Because you fall back to a poll mechanism which lets the application
deal with a larger chunk of work at once.

> the server is better off ignoring queue overflow.

If the net result of ignoring a signal is to completely freeze one or
more network connections, then I beg to differ.

--Stephen

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/