Firewalling MAC addresses
-
I notice the last response is 2mo old, but... I am using SonicWalls with MAC addresses being used in the rules all the time - for the very reasons that the OP listed, with IPv6 being the more annoying bit for me.
Filtering by IP is about as effective as milking a boar.
Since I've used pfSense for ages and the SonicWall bits are kinda new to me I had asked an engineer how the Sonicwall does this effectively with such an anemic CPU, and was told that the rules essentially are updated when the MAC goes live on the network with whatever IP(s) the machine has. The documentation I've read for the Sonicwalls I've been managing indicate that they operate very similarly to how pfSense does, aside from not being based on pf.
Since we have to record this information anyways and validate against DHCP/ARP, I'm not sure how-come pfSense can't do this? We're running CPUs that are worlds better than the junk in even the expensive Sonicwalls, so even if there was some overhead, I think it should be possible and even feasible. However I will digress in that I am not a programmer, and there may be some code sitting in the way that makes this incredibly expensive to do.
It sure as heck would make my life much easier. I could set things to happen based on MAC and not give a crap what IP they have or if they manually set it to something else, or if they're using IPv6 which as stated before by OP and here is ineffective.
-
All I can say is: Go Gary!!!!! You rock! Speaking from the year 2021 here, we have to deal with random macs now, so we should been on top of this problem when Gary first suggested we should! We “should have” had the ability by now to not give IP addresses to random Mac clients, just for example. OP GARY where you at???
-
@nick-wollman Why in the world are we resurrecting this thread...especially if you understand that MACs are randomized/spoofed...
-
@skogs good morning!
Honestly, I just wanted to congratulate Gary for pushing through all of the backlash he got. Wonderful, someone who knows what he wants and pushes to get it. I can relate, people just want to keep things the way they are. Or like to shut down honest question posters in the first reply.And now that I’m reminded, I just might have a solution that Gary is looking for, if he’s still around.
-
Rock on.
Deny by default works nicely. :) -
@chris4916 said in Firewalling MAC addresses:
What would be interesting, at least to me, is to understand why you would want to implement FW rules based on MAC address. This is something I don't understand yet.
easy answer : because devices may end up being assigned multiple IP addresses, a rule against the device's MAC address is a more "fixed" allocation
easy answer : because devices may end up being assigned multiple IP addresses, a rule against the device's MAC address is a more "fixed" allocation.
Plus it's nice to have options.
I am sure others would have additional use cases
Either way, it's a valid firewall parameter with the other big players (eg: CISCO, etc), so why not pfSense ?
-
@nick-wollman
.... waiting .... what's the solution you are proposing in pfSense --- I am all ears :-) -
IMHO MAC based fire walling is really really needed !! I did already advocate for that a couple of times.
Two reasons for that:
- given malware you do not only have to protect your network against threats from the internet, but also against threats from within your own network
- ipv6, at least from the firewall standpoint IMHO it is a disaster
So where you could filter in the IPV4 world on IP-address to limit access or do the opposite allow something based on IP, that option is gone with IPV6.
And in both cases IPV4 and IPV6 the MAC-address is a better option to filter a machine than one of its perhaps many IP's addresses and address variants.
Yes I do know that:
- it is possible to change the mac address
(it is even simpler to change your ip) and - that pf the firewall below pfsense regrettable does not support mac filtering
(please mail the maintainer on free bsd forums) - that it is true mac is level-2 not three 3
- that you can generate an alarm if a new mac occurs on your network (there is an app for that, if you think that is useful)
All in all nothing is perfect !, but mac-filtering is given IPV6, really really of big value.
Louis
-
@louis2 said in Firewalling MAC addresses:
malware
If the presence of 'malware' needs MAC handling more then IP, then, I agree, you have very untrusted clients.
Use the captive portal, which is meant to be used when you have to deal with non trusted devices, and you'll be using the firewall pf and the firewall ipfw. ipfw can do level 2 and 3.Still, I'd like to understand why the MAC, or even IP, is important when you deal with malware.
Malware could be found in the data payload, which is non accessible to you. Remember : there is no http traffic anymore. Or mail that doesn't use some TLS.
Also, you could see the destination IP, but not the destination MAC, as this would be the MAC of the next up stream router. -
@louis2 What's your environment / LAN type?
Are you talking guest type access here, if the devices only need to talk out their default gateway and not to other devices on the LAN then set the LAN ports to a protected port so they can only connect to the default gateway.
Run some sort of IDS/IPS on your LAN interfaces.
-
MAC filtering will do absolutely nothing for incoming packets. The only MAC address you will see is from the next upstream router and nothing else.
-
@louis2 said in Firewalling MAC addresses:
that option is gone with IPV6.
How so - I have specific rules that have specific IPv6 addresses in them.. I only allow access to my ntp server IPv6 address.
IPv6 clients can use multiple IPv6 addresses, if you do not tell them not too. But specific IPv6 addresses can be assigned to clients. Where they only use that address. But generally speaking with IPv6 you would do the rules based on any IPv6 address in that prefix. If you have some device that needs to do something different than the rules on that vlan, then move them to another vlan where you can set the rules you want for that prefix - now specific IP address doesn't matter.
While agree that IPv6 brings changes to how things were done in the past with firewall rules - filtering on mac address is not a requirement to correctly firewall.. You just need to understand the differences that IPv6 brings to the table. Yes there is a learning curve there for sure - which is why I suggest - if your not up to the task, or do not have time currently to climb up the learning curve for IPv6 and all the differences it has - then just don't use it. There is nothing saying you need to use or allow for IPv6 on your network. Many ISPs currently don't even provide it at all. Can you name 1 resource that requires you to use IPv6? If you do not have a pressing need for IPv6 - then you can put off climbing up the learning curve hill for a later date.. You have YEARS for sure ;)
-
I try to keep my computers patched and secure, but as you know, you can never be 100% sure that one of your computers has been compromised. So IMHO it is not wise to trust your own systems.
So, I try to do things to prevent "infected computers" to reach other computers. And the same rules would hopefully warm me that some thing strange is happening.
Capative portal does not help here, since that protect against unwanted computers, not against my own computers potentially being infected.
And of cause as I am talking about MAC-filtering, I am talking about traffic originated from my own vlan's
-
My remarks are related to multiple situations. Surely not only my guest lan!
I have multiple vlans and in each vlan I limmit the outgoing traffic, traffic to other vlans and/or the internet as far as possible.
Extreme example is e.g. the greenzone (e.g. my nas) which should be kept away from the internet as far as possible. Among other things I use floating rules to make sure that (nearly) every other vlan can not enter that vlan.
-
I know, mac filtering is (when supported) only possible for network internal generated traffic. And that is exactly what I had in mind
-
@louis2 how does mac filtering help in filtering intervlan traffic? Macs are not used between vlans.. IP and Ports and protocols are used between vlans.
If you have a compromised machine and it is allowed to talk to your nas on port 445.. (smb over tcp).. What does matter what its mac address is?
-
John, it is complicated not hew of course. A few remarks:
I only allow access to my ntp server IPv6 address
-
that alone is an issue but also in ipv4, problem is that a lot of applications use their own build in ntp, dns etc. So if you want to force the use of your own dns etc, you probably have to redirect the server address, which is as far as I know not yet possible for IPV6
-
Related to used addresses there are two different issues here
a) Different address types e.g. link local and global
b) Multiple and changing global addresses (fixed, temporarily, changing all the time for security reasons)
Additionally a server can have multiple addresses in favor of multiple functions / applications. So this all together makes it impossible to filter on IP-address.
What does to a certain extend helps is to create many vlans. Since that provides better separation and each vlan does have it own rule-set. More complicated to maintain of course and also not always a good idea. Reason for that is e.g. that the equipment in the separated vlans also have to talk to each other.
For info, I do have native IPV6 for years and there is one big advantage. I do have an endless number of IPV6-addresses, where I only have one IPV4.
My servers all have and use IPV6 next to IPV4. If applicable/needed they my servers are accessible from the internet via IPV6.
A lot of if not most equipment today has a preference for IPV6 over IPV4. So IPV6 is available it will use IPV6.
-
-
@louis2 said in Firewalling MAC addresses:
can have multiple addresses in favor of multiple functions / applications. So this all together makes it impossible to filter on IP-address.
While yes clients can have multiple IPv6 address - they don't have too.
If applicable/needed they my servers are accessible from the internet via IPV6.
Why? Do you have clients that only have IPv6? While again using IPv6, are you behind IPv4 nat and can only provide access that is not natted via IPv6?
So IPV6 is available it will use IPV6.
Which is another good reason that if your not up to speed on all the differences and changes that IPv6 brings to just not use it..
b) Multiple and changing global addresses (fixed, temporarily, changing all the time for security reasons)
Again they do not have too. My ntp server this is served to the public via ntp pool, its IPv6 it only has the 1 address I gave it - and it doesn't change.. It doesn't use temp IPv6 addresses, because I told it not too, etc..
If you are not ready to embrace all the changes that come with IPv6 - don't use it, or yeah your going to have to ramp up and learn how to handle the differences. Mac addresses are not the solution to firewalling IPv6 and as stated pfsense does not have any real support for using mac addresses in filtering other than captive portal, or limits by doing static arp, etc. If you need to or feel you need/want to use mac filtering in your network - then you prob better off using something else as your firewall that supports it. Pfsense has limited mac abilities from a firewall point of view, and I don't think they are going to be adding any new abilities in that area like next week ;)
-
Hi.
Sorry to revive this more than 2-years old thread.
I registered on this forum just to reply about the need for the ability of filtering / blocking by MAC address.
I am by no means versed in networking, however I have a basic understanding of the principles and learned my way around in PFSense because for the most part, I find it an excellent solution for my homelab environment.
I won't go into the ipv6 debate, I am not at all familiar with the subject. However, I am a bit privacy and security-oriented and have a small dose of healthy paranoia.
I'm also not sure if this is the right place to address this
, But here it goes, from a less technical user's perspective.
Here is a practical and basic scenario where this would be useful, :- Let's say you have your simple home network having both LAN and WLAN. Let's say someone has managed to crack your WiFi password. This means he now has unrestricted access (unless your WiFi AP has the ability to prevent it) to your entire network. If you have the ability to set rules by MAC Address, then you can either block any network access for that unknown MAC by denying an IP, or you can kick it into an isolated VLAN. In order to spoof a MAC on the intruder device, the intruder would have to make the extra effort of finding a MAC address which is allowed on your infrastructure and even so, there could be a feature which raises an alert and prompts for manual intervention if a duplicate MAC address is detected in your infrastructure. This would mean that even if somehow the intruder KNOWS your IP ranges and sets their IP manually, they would still be blocked from accessing the infrastructure.
- Regarding the new random MAC feature on Android devices: In the same context of a small home network, if you trust an Android device, then you should also have that Android device trust your network. And the same modern Android devices give you the option to use the real MAC address for your particular network, therefore also successfully integrating into this concept. This will make a new connection a bit more complex, but if you're willing to go that route, you will also take this step.
- A first additional layer of security which could be added would be an OS fingerprint feature on top of that. So if your known MAC address is an Android device, but the intruder uses a different OS, even if they manage to spoof the MAC address, it will kick them out. Let's say that they try emulating other OS fingerprints until they find the correct ones. If several such OS fingerprint changes are detected, then the device still stays kicked out.
- A second additional security layer would be that, besides the fingerprinting, you could have some simple cross-platform client, which after connecting to the network, also provides some security key., known only to the firewall. This could also complement the already existing 802.1x security features if your switches support that, but also allow for better security for those of us who don't have the resources to invest in equipment with more advanced security.
Thus, you would be adding some additional control and complexity to even the simplest network setup, which will not necessarily block any and all attempts at infiltration, but definitely discourage an intruder by adding more layers to hack, therefore making it more difficult and time-consuming to infiltrate. This will, however, most likely stop the neighbour who was smart enough to find some tricks to hack into a WiFi and forgot to pay his internet provider and wants to use your internet.
On a small-scale home network, even in a homelab context, as these additional security featured would be implemented once for a particular device, they can be done manually.
I am sure that at an enterprise level, these can be centralised somehow.Rant: Sorry if this may rub someone the wrong way, but if a project relies on its community for feedback and for useful feature requests which can be later incorporated into the paid version, which targets the corporate environment, features which are already being provided by other suppliers, but then it refuses those suggestions because they cannot be implemented quickly is not a particularly healthy attitude. I can see this also in other such mixed projects, like Nextcloud, but that's another story.
/rant.As mentioned or hinted, I am not exceptionally technical and knowledgeable in this context. Perhaps I am not using the proper terms, but I hope that this post helps others who understand the technical parts better to drive the implementation of such features which nowadays should be considered common sense...
-
Since pfSense can only filter packets passing through it, it can do nothing to keep those MAC addresses off your local network. About the best you can do is only provide DHCP addresses to known MAC addresses.
