Re: TCP RST handling in 6.0

2005-11-09 Thread Mike Silbersack
On Tue, 8 Nov 2005, Lars Eggert wrote: Also note that other attacks against long-lived TCP connections are still possible, e.g., through spoofed ICMP packets. I don't think we've been vulnerable to the ICMP-based reset attack for a few years, actually. Using SYN packets is the best method,

Re: TCP RST handling in 6.0

2005-11-08 Thread Marc Olzheim
On Tue, Nov 08, 2005 at 01:56:41PM -0800, Lars Eggert wrote: > On Nov 8, 2005, at 12:46, Marc Olzheim wrote: > >Being on the wrong end of a distributed tcp syn flood attack atm. > >on the > >machine I'm mailing from, is probably enough to convince me of its > >use. > > The change we are discus

Re: TCP RST handling in 6.0

2005-11-08 Thread Lars Eggert
On Nov 8, 2005, at 12:46, Marc Olzheim wrote: Being on the wrong end of a distributed tcp syn flood attack atm. on the machine I'm mailing from, is probably enough to convince me of its use. The change we are discussing is not protecting you from SYN floods, it is supposed to protect you f

Re: TCP RST handling in 6.0

2005-11-08 Thread Lars Eggert
On Nov 8, 2005, at 11:54, Mathieu CHATEAU wrote: 1/it can be set back if needed It can be enabled, too, if needed. 2/95% of users will get benefits against 5% that will disable it I'd love to see a source for those numbers. 3/over the time, i am having above 70 lines in sysctl.conf to get

Re: TCP RST handling in 6.0

2005-11-08 Thread Lars Eggert
Hi, On Nov 8, 2005, at 11:23, Mike Silbersack wrote: I'm open to discussing the change. I plan to revisit that and the SYN causing a connection reset issue after eurobsdcon. good to know, thanks! However, I'm open to clubbing you over the head for not saying anything throughout the enti

Re: TCP RST handling in 6.0

2005-11-08 Thread Marc Olzheim
On Tue, Nov 08, 2005 at 11:02:25AM -0800, Lars Eggert wrote: > Thus, I'd like to suggest that the default for > net.inet.tcp.insecure_rst be zero for now. AFAIK, any other TCP mod > came disabled be default in the past, too. Being on the wrong end of a distributed tcp syn flood attack atm. on

Re: TCP RST handling in 6.0

2005-11-08 Thread Brooks Davis
On Tue, Nov 08, 2005 at 11:02:25AM -0800, Lars Eggert wrote: > Hi, > > I came across the following in the release notes of 6.0 recently: > > "The RST handling of the FreeBSD TCP stack has been improved to make > reset attacks as difficult as possible while maintaining > compatibility with the

Re: TCP RST handling in 6.0

2005-11-08 Thread Mathieu CHATEAU
hello, to start with, i don't want to raise a troll... argue to keep it set: 1/it can be set back if needed 2/95% of users will get benefits against 5% that will disable it 3/over the time, i am having above 70 lines in sysctl.conf to get FreeBSD secured and the network strong and fast. 4/the 5%

Re: TCP RST handling in 6.0

2005-11-08 Thread Mike Silbersack
On Tue, 8 Nov 2005, Lars Eggert wrote: Thus, I'd like to suggest that the default for net.inet.tcp.insecure_rst be zero for now. AFAIK, any other TCP mod came disabled be default in the past, too. Lars I'm open to discussing the change. I plan to revisit that and the SYN causing a connec

TCP RST handling in 6.0

2005-11-08 Thread Lars Eggert
Hi, I came across the following in the release notes of 6.0 recently: "The RST handling of the FreeBSD TCP stack has been improved to make reset attacks as difficult as possible while maintaining compatibility with the widest range of TCP stacks. (...) Note that this behavior technically v