Nelson Minar <[EMAIL PROTECTED]> wrote:
>
> If one side of the TCP connection is still working, will a keep-alive
> from the working side eventually clear out the stale TCP connection?
> If so, then it seems like the IP-Masq default timeout for TCP should
> be above whatever the common keep-alive timing is.
Seems that way to me, too. But keep-alives are an OS-dependent thing.
Does Windows do keep-alives? Does NT? Does MacOS? What values do they
use for their timeouts? Who knows?
> Anyway, keep-alives could still be useful for keeping the connection
> "fresh" in IP-Masq's eyes. Except they're off by default.
I know that ssh, for one, enables keep-alives for its connections, but
as you noted, the keep-alive timeout is far above the masq timeout. And
that's only one OS, Linux, that we're thinking about. Windows may do
something completely different.
Keep-alives can't be relied upon, so the masq code institutes its own
timeouts. That's the short story.
Since the timeout can be changed, go ahead and figure out what your
needs are,and change the timeout to whatever is reasonable for your
site.
--
[EMAIL PROTECTED] (Fuzzy Fox) || "Her lips said 'No,' but her
sometimes known as David DeSimone || eyes said 'Read my lips!'"
http://www.dallas.net/~fox/ ||
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
For daily digest info, email [EMAIL PROTECTED]
