Hello friends,
How can I distribute three public IP via PPPoE? But I do not have a whole subnet to public IP, only a part. My idea was to allocate the three clients via PPPoE a private IP, and then assign each fixed to public IP NAT (no masquarade). Both IPs are 1:1 translated.
PPPoE-LAN <==> WAN
10.0.0.1 <==> 77.88.99.111
10.0.0.2 <==> 77.88.99.112
10.0.0.3 <==> 77.88.99.113
PPPoE-LAN works fine, but I search only for a NAT-Rule which translate that for- and backward.
How can I do that?
mikrotik75
Proxy ARP:
http://forum.mikrotik.com//viewtopic.php?p=2010 [just one of many threads about this]
Thank you for your reply. I don’t understand this, but I’ve found another site in wiki:
http://wiki.mikrotik.com/wiki/How_to_link_Public_addresses_to_Local_ones
This way I’ve understand and it works great.
mikrotik75
OK, re-read your post, you were asking for 1:1 NAT
Well, if your customers start demanding actual public IPs, you know where to start anyway.
This night I’ve tested my configuration on our real system. Now I know what you mean:
This is functionally:
- client internet connection via PPPoE
- public IP for specific LAN-client via NAT (LAN=private IP <==> WAN=public IP)
Problems:
- the LAN-client get only a private IP like 10.0.0.1 via PPPoE, but on WAN-site a public IP like 77.0.0.1
- the LAN-client should have the public IP direct (without NAT), because IPSec makes problems
How I can give the PPPoE-LAN-clients a public IP direct? I mean how I can route it?
Thanks
mikrotik75
Hhmm, moment, I should read the lines careful. I try to understand the thread…
I’ve activate “proxy-arp” on eth1 (WAN) and eth2 (LAN) and deactivated NAT. It looks like good. The first tests on my dummy-system works fine. Should it really so easy?
This night I can test in our real system and I would see if also works fine…
mikrotik75
This night I tested the config on our real-ISP-system again.
Unfortunately this config works not on our real-system. I don’t know why. Now I have built a better dummy-system with the same environment-IPs like the real-ISP-system.
What I have done?
- set interface ether1 (WAN 77.x.x.192/27) to proxy-arp
- set interface ether2 (LAN 10.0.0.0/24) to proxy-arp
- disable any NAT rules
For me it is not clear whether the PPPoE-client is assigned a private or public? If the client become a private IP, I should route the public to the private IP? Is it not the same like NAT?
My problem is the topic-translation to german. I think I’ve understand some points wrong of instructions.
A example source code would be great.
Thanks for all help!
mikrotik75
P.S.: I given up first before. I do not know how to continue. 
Now I’ve found this: http://wiki.mikrotik.com/wiki/OSPF_and_PPPoE_Setup.
With this configuration I can give the LAN-clients a real public IP. And the LAN-clients can get internet. The outgoing connection worked good. But a IPSec-connection (from PPPoE-LAN-client with public IP to a remote router) runs not really. The time by ipsec / remote peer runs (initiator) and the peer-router says also connected, but the connection time on peer there is not run. Also no data transfer is possible.
I think there is no lack much. Maybe a firewall rule or route entry?
I am grateful for any information!
Merry Christmas
mikrotik75
If the clients get internet access then it’s probably not a routing issue. Is there any NAT involved? When a client browses to the web, what public IP do they get? If it’s not NAT or routing then perhaps it’s a firewall rule.
The clients gets internet with the right ip. NAT is not used there. I should search for a right firewall rule. I think I need a forward rule from WAN to LAN interface (backward rule
).
Thanks for your reply
mikrotik75
Just I have seen that still a NAT rule was working for the public IP pool and the public IP was not the right one. Browsing are possible, but with the natted IP address (wrong IP). I guess, thats why the IPSec transfers fails.
Now I’ve deactivated the NAT masquerade rule. Now the client have no internet connection. The LAN-client get via PPPoE the right public IP, but only with the OSPF-routing the thing isn’t running. The LAN-client with the public IP isn’t visible on the WAN-site. What can I do now?
Thanks for all reply.
mikrotik75
I’ve found out, how to get a working PPPoE connection with public IPs. The trick is an own bridge with own public IP of ISP-range for the WAN interface. The bridge is needed for the PPPoE-Profile. The router use then two public IPs for working. One for NATted-connections and one for PPPoE-connections with public IPs.
Only a small mistake the system have:
If I like to use the connection with a public IP via PPPoE, I need a initial-connection with an local IP via PPPoE (NAT). Then I can close this connection and the new connection with public IP over PPPoE runs fine. I think the system is needed an established connection, but why?
I think it’s only a little step by firewall rules. Have someone an idea?
mikrotik75