[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <24950.1457471469@warthog.procyon.org.uk>
Date: Tue, 08 Mar 2016 21:11:09 +0000
From: David Howells <dhowells@...hat.com>
To: David Miller <davem@...emloft.net>
Cc: dhowells@...hat.com, linux-afs@...ts.infradead.org,
netdev@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 01/11] rxrpc: Add a common object cache
David Miller <davem@...emloft.net> wrote:
> > One could argue, I suppose, that things should've been arranged that the
> > RxRPC client would manage the lifetime of each connection it sets up,
> > rather than both ends letting it lapse by mutual loss of interest. But
> > you *still* have to have a timeout, lest the client die and not close its
> > connection.
>
> But the point is if there is no limitation on the number of these connections
> that can be setup.... wait for troubles.
>
> And if you try to put in some kind of limit to handle this, it's then
> easy for the bad guy to block out other legitimate users.
I can put in a limit per peer, where a 'peer' is either a particular remote
UDP port or particulat remote host. TCP has this by virtue of having a
limited number of ports available per IP address. But if I have 10 IP
addresses available, I can attempt to set up half a million TCP connections to
a server simultaneously. If I have access to a box that has an NFS mount on
it, I can potentially open sufficient TCP ports that the NFS mount can't make
a connection if it's not allowed to use privileged ports.
Am I right in thinking that you have decided that it can't be done and
shouldn't be done?
David
Powered by blists - more mailing lists