Well, I can’t be sure of the wan up address without being at the site (I believe).So the dyndns address check out to the current WANIP of the remote device and you can ping the device but WG does not come up??
Did you make any changes to the config prior to losing connectivity as there is no clear reason I can think of that would cause loss of connectivity.
Only point was EoIP is easily encrypted with the "Use IPSec" checkbox, which uses a PSK defined on EoIP interface... which is kinda handy. Since pure IPSec is a lot of config...EOIP works between two IP addresses and doesn't care about how its packets move from point A to point B. So one can use any kind of connectivity to do the job. Since EOIP doesn't do any encryption, it's wise to use something that does it. IPsec is fine, wireguard is fine, etc.
Yeah but without certificate how safe is it.................
The idea is EoIP just carries RoMON protocol. If EoIP is NOT bridge to anything, and each end has a unique IP address in same subnet (and subnet it NOT used by anything else), there shouldn't be much traffic since nothing route to it. I would NOT use EoIP to replace your existing WG – use that for normal traffic. The idea here is to have some "backup" management interface beyond WG (e.g. in case you misconfigured WG remotely).I like the idea of EoIP because of the advantages of layer 2 connectivity to all devices, but I am concerned about traffic or taxing the MT devices with firewall rules.
Wouldn't this require bridge between eoip and etherX on remote device?So if the EoIP terminated at some central router, it be able to see anything with RoMON enabled – even if it's two hops aways (e.g. hub router --(eoip)--> remote --(etherX)--> ap).
That is a great way to conceptualize the use of eoip, and a great way to make an initial (and possibly final) implementation.The idea is EoIP just carries RoMON protocol. If EoIP is NOT bridge to anything, and each end has a unique IP address in same subnet (and subnet it NOT used by anything else), there shouldn't be much traffic since nothing route to it. I would NOT use EoIP to replace your existing WG – use that for normal traffic. The idea here is to have some "backup" management interface beyond WG (e.g. in case you misconfigured WG remotely).I like the idea of EoIP because of the advantages of layer 2 connectivity to all devices, but I am concerned about traffic or taxing the MT devices with firewall rules.
The trick is since EoIP is ethernet-like interface, it works with RoMON. And since RoMON on a local router will find more routers on other "real" ethernet lines. So if the EoIP terminated at some central router, it be able to see anything with RoMON enabled – even if it's two hops aways (e.g. hub router --(eoip)--> remote --(etherX)--> ap).
There is nothing special to configure on RoMON to use EoIP, other making sure RoMON is enabled and secrets match. You can limit the interface RoMON will use, but obviously it have to include at least the EoIP interface (under ports).
True. It's actually setting the "IPSec Secret" that enables in V7 – I'm remembering V6 I think with checkbox.Dont forget to tick the ipsec security otherwise you are creating an open hole at both ends.....
Which is okay, you're really just using it for management. But all encryption has to flow through the CPU anyway.Note that when I added an ipsec key, I was forced to turn off "Allow Fast Path"
Here was the post: viewtopic.php?t=203951&hilit=eoip+nat#p1053229Id be interested in your only one side needs public IP teaser. Please elaborate!!!
Just tested with Fiber+Static to LTE+CGNAT with EoIP+IPSec: It works. But still need use the CGNAT'ed remote address on the "static IP" side. But you can use the DDNS name of the CGNAT side in EoIP config (even though it says "Behind a NAT" in /ip/cloud) for that. No need to set a local address in EoIP (in fact with IPSec, it just be more confusing) & tunnel ID's need to match – but that's it. Beyond enabling RoMON.Id be interested in your only one side needs public IP teaser. Please elaborate!!!
Makes perfect sense -- thanks.Which is okay, you're really just using it for management. But all encryption has to flow through the CPU anyway.Note that when I added an ipsec key, I was forced to turn off "Allow Fast Path"
I wrote an article here explaining:AMMO, can you test with IPIP instead or at least tell me how to do so beyond the standard settings.
aka. for client site with no static public IP --> what do I put for local address?>>
aka. for server site with public IP --> what do I put for remote address??
What additional firewall rules are required since I am not picking any particular ispec just a secret are there are any ports to open on input chain for server device.
Any funky source nat rules etc...................
/ip firewall address-list
add address=other-router-DDNS.sn.mynetname.net list=gre-allowed
/ip firewall filter
add action=accept chain=input comment="allow IPsec NAT" dst-port=4500 protocol=udp
add action=accept chain=input comment="allow IKE" dst-port=500 protocol=udp
# [...]
add action=accept chain=input comment="accept GRE" protocol=gre src-address-list=gre-allowed
# [...]
add action=accept chain=forward comment="defconf: accept in ipsec policy" ipsec-policy=in,ipsec
add action=accept chain=forward comment="defconf: accept out ipsec policy" ipsec-policy=out,ipsec
/ip firewall nat
add action=masquerade chain=srcnat comment="defconf: masquerade" ipsec-policy=out,none out-interface-list=WAN