Hi Kevin,

On 20/12/16 01:10, Kevin Long wrote:
> I was just browsing the Mastering OpenVPN book and a paragraph jumped out at 
> me which basically said that using OpenVPN on port 443 is a common way people 
> try to duck firewalls.  Indeed, this is what I do.  My clients are all over 
> the place, airports, hotels, different countries etc, and we do seem to have 
> better luck on port 443 tcp than 1194 tcp or udp.
>
> But the book states, as I have just learned just recently coincidentally,  
> that OpenVPN traffic (even running on TCP) does not really look like normal 
> browser TLS traffic.
>
>
> I saw in the release notes I believe, that the new tls-crypt feature helps 
> prevent metadata about auth certificates from being exposed, as well as 
> blocking deep-packet inspections of the traffic.
>
> Could anyone possibly elaborate on this? Will this in practice help do 
> mitigate OpenVPN blocking on port 443 in cases where normal TLS 443 traffic 
> is permitted?
as author of that part of the book, I feel obliged to respond ;)

the new tls-crypt feature adds some extra hiding of OpenVPN traffic from 
someone snooping the network. It does *NOT*, however, result in OpenVPN 
traffic looking the same as regular TLS/HTTPS traffic. If you scan an 
OpenVPN network connection using tcpdump/wireshark then you can see that 
it's different from a regular HTTPS connection start. Advanced firewalls 
can detect this difference and can/will block OpenVPN traffic based on 
it. Currently, there's little that can be done about this.

> Also, could anyone elaborate on tis-crypt being “poor man’s quantum” 
> protection
>
>
when/if a working quantum computer is available then the current key 
exchange mechanisms such as RSA, DSA and even Elliptic Curves no longer 
offer protection: a quantum computer can supposedly break this key 
exchange within seconds. A quantum computer cannot break encryption 
itself (supposedly), only the initial ephemeral key exchange. With those 
keys, however, encryption can also be broken. By adding something like 
'tls-crypt' we add an extra preshared secret, which a quantum computer 
cannot break so easily, thereby adding what we like to call "poor man's 
quantum protection". Whether it holds in the future (or is even 
effective at all) remains to be see when quantum computers and quantum 
computer algorithms develop.

HTH,

JJK



------------------------------------------------------------------------------
Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today.http://sdm.link/intel
_______________________________________________
Openvpn-users mailing list
Openvpn-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openvpn-users

Reply via email to