Re: Proxies, opensource and the general market: what's wrong with us?
Claudio Telmon <[email protected]> Wed, 27 Apr 2011 23:59:52 +0200
| Newsgroups | gmane.comp.security.firewalls.wizards |
|---|---|
| Message-ID | <[email protected]> |
On 04/27/2011 10:52 PM, David Lang wrote: > however, as proxy firewalls are dieing, new devices with the type of > checking that proxies do are becoming more common. > I don't think so. No product that I'm aware of has the same "default deny" on the low level attacks that a proxy has. Again, the recent "split handshake" problems are a clear example: packet filters "try to guess" the proper session state, while there is no way to cheat a proxy into letting a connection in if it's not permitted (up to TCP/UDP, I mean). Packet-handling tools, be it filters, IDS or something else, however, are probably "good enough" for the market. > doing the checking with a proxy listening to a specific port should be > significantly easier thatn checking for all protocols on all connections > passing through the devices. > It is, actually, if it's TCP. For what I remember as I wrote some code in this area, UDP is much more of a nightmare. This is why I say that proxies are good for some protocols (e.g. http) where you can benefit from tight controls, but you still need a packet filter underneath for other protocols: you can't punch a hole in a proxy for a new, unknown and "essential" protocol. ciao - Claudio -- Claudio Telmon [email protected] http://www.telmon.org