Coming from zyxel its not too bad, no loopback checkbox and little less spoon feeding. (used to be a reseller for zyxel)
The main difference is only one firewall rule for all port forwardings is done.
The main place for rule details are in the dst-nat rules.........
You have the right info.
You have two choices in Mikrotik for formatting the Dst NAT RULES
TYPICALLY
a. for dynamic firewall rules one uses in-interface=ether1 or in-interface-list=WAN (assuming ether1 is the WAN interface, like standard cable)
b. for static, fixed ip, one uses dst-address=PublicIP of WAN
If you think about they say different things but have the same result. The first says for any traffic coming from the WAN side, headed for portxx send the traffic to lanip blah blah
THe second one says that for any traffic heading to my public IP, headed for portxx send the traffic to lanip blah blah.
Pretty straightforward !!
Note: You could use the dst-address for dynamic IPs, but then would have to change the rule everytime the IP changed.
The in-interface list method, for Dynamic WANIPs, is useless if you ever want users or yourself (FOLKS ON THE LAN behind the router) to access the server via the Public WANIP.
I never understood loopback on the zyxel until I ran across this need. I was like, pffffft what losers dont just use the LANIP directly of the server when within the LAN. Clearly there are many cases where people seem to defy logic.
So there are work arounds to make the dynamic IP scenario LOOK LIKE the fixed IP scenario, so that the router will accept both internal and external users access to the server.
One can see right away that if you use in-interface-list=WAN one is NOT INCLUDING internal users.
However as noted above I could use the IP address but would have to change the rule everytime the IP changed, so how to avoid that is the next format!!
FORMAT
add chain=dstnat action=dst-nat dst-address-list=(applicable firewall address list for your dyndns of choice - most use the IP cloud address on the router itself)
dst-port=xxxx protocol=tcp to-addresses=IPofServer {to ports is not required if the same as dst-port}
What we have done here is to identify the public dynamic IP with a dyndns name which will keep the IP relatively current and thus one can use the format of the fixed IP dst nat rule which would include both internal and external users because they are both pointing at the public IP.......
There is one caveat, and that is if the server and the users are in the same subnet we run into something called hairpin NAT. All explained in the article below and more.
Suffice to say one source nat rule fixes this issue, and alternatives are discussed as well.
Enjoy todays instructional lessons!
viewtopic.php?t=179343
I should note that the dstnat rules can also have source addresses or source address lists which narrow down access to the server, always a good thing!
Since people can get free dyndns names for their own routers, they can give you their static fixed wanip or dynamic URL to put into the firewall address list!!