Subject: Re: several messages (about SOMAXCONN)
To: Marc Slemko <marcs@znep.com>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: tech-net
Date: 09/14/1998 19:49:22
>That doesn't work because of history. History says that SOMAXCONN is
>a reasonable value for an "average" app to use. If you make SOMAXCONN
>INT_MAX, you break that.
I don't buy that. Your previous post shows that "historical" usage is
broken anyway. How does a range of 3 orders of magnitude count as a
"reasonable" for an "average" app? If you're seriously saying that a
backlog of 5 is comparable to a backlog of 1000, then I for one don't
take the argument seriously anymore.
That woudn't mean you're not 100% right, tho'.