httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jim Jagielski <>
Subject Re: Lingering close: a summary
Date Mon, 10 Feb 1997 00:01:20 GMT
Marc Slemko wrote:
> I'm not sure you can say TCP is broken.  After all, lingering_close() does
> exactly what we want.  If you want it at the API level, SO_LINGER working
> as it should does it too.  TCP != the API.  The protocol itself isn't
> necessarily broken.

SO_LINGER will, if properly implemented, close the link after a period
of time... Thus it is entirely possible that it will be closed
before the client is aware of what is going on, leaving us in the
same situation as before. We can hand wave and say that the client should
be aware after this amount of time, but we can't guarantee it. Thus,
carrying things to the extreme, Apache should be willing to wait
forever because it "needs" to see that the client is aware that
the link is closing and it also needs to "know" that the client
is "definately" not going to send anymore data...

I agree that l_c() does what we want; it also does some nasty things
we don't :)

      Jim Jagielski            |       jaguNET Access Services           |
                  "Not the Craw... the CRAW!"

View raw message