The DHCP log on my pfsense gateway is showing a large number of failed IPv4 DHCP unicast RENEW attempts. For some reason it seems that the Rogers DHCP server (220.127.116.11) is returning DHCPACK packets with the dhcp_server_identifier option populated with the same IP address as the bootp relay agent option (18.104.22.168). That relay agent IP does not seem to be responding to my unicast RENEW requests.
Aren't uinicast RENEW requests supposed to bypass the relay agent and go directly to the DHCP server? If so shouldn't the dhcp_server_identifier option specify the IP of the DHCP server, not the relay?
Eventually (after many hours and failed attempts), my gateway gives up on renewing via unicast and succeeds by sending a broadcast RENEW attempt.
Is there something in my client config that could be causing this? This started about a month ago and I didn't make any changes on my side.
Thanks for gracing us with your presence in the Rogers Community Forums and nice first post! 😁
Internet performance issues can certainly be a pain to deal with, especially if it seems to have popped up out of nowhere! I know you mentioned that you hadn't made any changes to your set-up prior to the issues occurring. Are you presently using any third-party equipment with your modem? Do you find that your internet is just slow or does the signal drop out completely?
If possible, please post a log of your signal levels making sure to leave out any private IPs.
We can also run some tests to your modem at our end but will need you to send us a Private Message @CommunityHelps to get the ball rolling. For more information on how our Private Messaging system works, please CLICK HERE.
Perhaps our Resident Expert @Datalink may have some insight to share on this topic as well!
Hi Laura; sorry to see they've got you working on a holiday weekend ;(
As I mentioned, my gateway is eventually able to renew its IPv4 lease via a broadcast request (at the rebind deadline long before it expires), so this is not having any negative impact on my service. It's just causing my gateway to send a few hundred futile renewal requests to what appears to be the wrong IP address every time my lease comes up for renewal. The only minor negative impact is that all these failed renewal packets create uneccesary traffic on the Rogers network.
My gateway is a Netgate SG-1100 running pfsense, my modem is running in bridge mode. This is why I'm able to see this happening in my logs.
I've started a packet capture for you so you can see what's going on, but I need to let it run overnight since my current lease is not due for renewal until 3AM. I can PM the capture to you tomorrow.
@harrdou the Rogers staff member who will have an interest in your observations is @RogersSergio. He will pass along your observations to the Rogers engineering staff or to Hitron's engineering staff if required. Just to note, @RogersSergio hasn't been seen in the forum in many weeks, as he and the engineering staff are preparing for the next firmware release for the CODA-4582 modem.
Unfortunately there is no point in communicating your observations to the tech support staff or moderators in the forum as this is a community based help forum, not a full blown tech support forum. This issue is well above this forum and Tech Support, unless of course the techs have access to a setting that can be easily changed (probably not), so, the engineering staff are the best staff members to look at this issue.
Run the data capture so that you have it on hand. I'll probably ask you to forward that to Sergio unless Sergio does that himself.
I'll send Sergio a message to have a look at this thread, however, I don't know if or when he'll respond to it.
Done ....... message sent.
As @Datalink has said we will need the Rogers engineering team to take a look.
I have contacted them and will work internally to get assistance to be able to respond to your situation.
I don't have an ETA at the moment but will continue to work internally to provide direction as soon as I can.
If you do have any additional info please let me know so we can get it over to the engineering team