Issue with huge numbers of connections

[Robert Watson]

On Sun, 17 Jun 2007, M. Warner Losh wrote:

> I have a friend who is having problems with a service he's running. He gets 
> billions and billions of connections to this service a day. Somewhere 
> between 10^8 and 10^9 connections, he notices that his servers lose the 
> ability to accept new connections.  These are TCP connections.
>
> This is with FreeBSD 6.1R.  My first question is: does anybody know if the 
> fixes to -current/7.0 have fixed this?  Is there a fix that can be back 
> ported?  He's currently working around the problem by having a number of 
> different machines that reboot in a round robin fashion, but would like a 
> better solution.

I think we'll need some more detailed information to make headway.  What does 
"Lose the ability to accept new connections" mean, and how firmly is it meant? 
Are we talking about "If you restart the application, it comes back" or 
"Reboot reuiqred", "SYNs don't get SYN/ACK's" or "SYNs get RSTs", or what 
exactly?  There's a lot of scope for exploration, and plenty of bug fixes in 
the tree, but it's hard to say whether any apply without a bit more detail.

Robert N M Watson
Computer Laboratory
University of Cambridge