Bug #6524
closed192.168.100.0/24 subnet with Cable Modem WAN un-workable
0%
Description
So here is the scenario i ran into today.
Customer Subnet is 192.168.100.1/24
Customer WAN IP is 174.77.55.XX6/28
Customer GW IP is 174.77.55.XX5
If the cable modem on the WAN side is of a model which has a status/diagnostics page on an ip of 192.168.100.1 as the following models do: Motorola, Scientific Atlanta, Cisco, and numerous others, The WAN becomes completely un-usable in this situation, occasionally a packet or two goes through, but 99% of the traffic is returned by the gateway with Destination Host Unreachable.
The only solution was to re-IP the whole customer network to 192.168.10.1/24 which took me an hour plus to do. This used to work on older versions as some time back i used a simular set of subnetting on a different site with an older 2.2.X build of pfsense.
I noticed on the console while watching PFsenses WAN interface attempt to DHCP when i removed the static IP, it was getting random IPs in the 192.168.100.0/24 range, but it would release the ip almost in the same second as it would get it.
This also caused a condition where the PHP-FPM would hangup or such and i wouldnt be able to get to the web admin, or it would lag really really bad, and restarting the FPM resolved that whenever it happenned.
I know that its unlikely to ever encounter such a scenario but it might behoove to note this behavior and advise against using such a subnet, or else fix the cause of this erroneous behavior.
This was tested on the latest 2.3.1 P5 firmware as of todays date.