06-27-2014 01:49 AM
I have an EdgeMax Lite configured with DHCP WAN on eth0 and a bonded LAN on eth1 and eth2.
I configured this using the config file here:
I then made a few changes, mostly to put the router on the 192.168.0.0/24 subnet and setup a new user and password. I also enabled upnp using the wizard.
The WAN connection is provided by a Virgin Media 'Superhub' configured in 'Modem Mode'. I dont beleive the details of this are important (but I am happy to supply anything that is required), as it is effectively a DHCP ethernet connection to the internet with no NAT or similar running on the modem. eth0 is given a external IP address on the Virgin network.
I set the system up last night and it was all working very well. When I got up this morning, there was no connection to the internet. I did a bit of digging and found that the problem seemed to be between the Edgemax Lite and the VM 'Superhub'. I disabled and re-enabled eth0 and the internet connection started working again.
I am assuming that this will be a regular occurance rather than a one off, but even still I was hoping someone could offer some guidance on how to debug the issue? If it does happen again is there anything that could tell me what has caused the issue? Can I enable any additional logging?
Is there anything additional I could, or should, set up in order to use the router with such a DHCP WAN interface?
Many thanks in advance for any guidance or tips.
06-27-2014 07:07 AM
I have had similar problems with the 2Wire router provided by AT&T U-Verse in the past. The problem was that 2Wire hardware not my gear. Soetimes it would work for weeks sometimes it died in less than a day. Restarting the WAN port on the router always let it pull a new IP from the ISP.
06-27-2014 11:12 PM
Thanks for the reply.
I strongly suspect it is the model at fault and not the ERL. I also thought it may be a DHCP address change that the modem does not handle correctly.
I believe my best bet would be to run a cron job on the ERL to check that there is working WAN connection and cycle the port if not.
Could anyone confirm whether there is a best practice method of doing this or should i simply add a bash script as a cron job to: ping a known IP or IPs, check the result and issue an "ifconfig eth0 down && ifconfig eth0 up" style reset?