Hi,
I'd like to ask a general question about what UDP connections. Please forgive me if my post doesn't make much sense - I'm not a pro! :o
In case of TCP connections, I understand that enabling an OUT connection basically means that the application will be able to establish an outbound connection, while granting IN access allows remote hosts to connect to the given port, provided the application in question is listening on it.
But from the moment the connection is actually established, data can flow in both directions (in and out) without limitation.
Now, since UDP is "connectionless" there's obviously a difference in the way the IN and OUT restrictions are interpreted. I mean, since there's no connection here, what does IN and OUT in a UDP rule actually mean?
Say I have a DNS request. AFAIK this works like this: the client application sends a UDP packet on port 53, and then, eventually, the answer also arrives by means of a UDP packet. Now, does that mean that every application that wants to make DNS queries (i.e. act as "DNS client") technically needs to have a rule with both IN and OUT traffic enabled??
Another example is DHCP. This is even more confusing as the initial request is sent to a broadcast address...
Could someone who understands this stuff please try to explain to me what does actually take place inside the firewall if I e.g. create a rule that enables IN (but disables OUT) traffic on a given UDP port?
Thanks
Hugo
Hi Huho,
welcome to the forums! Hopefully some info here (http://outpostfirewall.com/forum/showpost.php?p=42666&postcount=19) can help clarify things.
 I could use some help again...
 Lost Licences |
What does IN and OUT actually mean for UDP connections??