NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Solved
NBN issue with latest firmware versions
I use a RBR750 with 2 satellites and they were just running perfectly. I am connected to NBN and I never needed to reboot and the connection to the internet was stable and great. Later came firmw...
It has been pointed out from the logs that there are many DHCP requests from several sources followed by a disconnect event. Thank you for this!
I now have assigned most of my devices a reserved IP address and have also enabled IPv6 on the ORBI.
Then I updated the firmware to the latest version 7.2.6.31 which never worked longer than 1 day. For the FW upgrade, I used this time the ORBI app on my phone.
Since then, no more crashes, everything runs very smoothly.
I will further monitor it, but it seems to me these frequent DHCP requests have caused the system to crash. Strange though, a DHCP request should be a very basic function...
"they found out that the modem (my Netgear RBR750) caused this issue." The RBR750 is not a modem nor does it have a built in modem. This is a router unit.
What brand and model is the NBN modem or ONT that the RBR is connected too?
Be sure your using a good quality LAN cable between the modem and router. CAT6A STP is recommended.
Does the ISP support DHCP WAN configuration or something different like PPPoE?
After v.21 was loaded, did you factory reset and setup from scratch with out loading any backup configurations from file soon after the system was updated?
You’re right, the RBR750 is a router…
The Modem is from NBN and belongs to NBN, this is part of the infrastructure here. The model is G-240G-P and details can be seen here: https://apcoms.com.au/nbn-fttp-connection-box-status-lights/
From this NBN modem, I have a CAT6E cable in the walls to the centre of my house which is about 15m long. I have replaced this cable 2 times with 2 new temporary cables, but this made no difference.
My ISP supports DHCP WAN, it does not require a login and it is not PPPoE.
I did a full factory reset after v 7.2.6.21 and configured it all from scratch.
My ISP did some investigation and they said my router is causing this issue, I should try with another router. But after rolling it back to v 4.6.14.3 it just keeps working fine for months without rebooting.
With any later FW version I get this issue daily, so the only logical explanation is that the firmware causes this.
What a dismal situation. It would be useful to know one item of network trivia: what is the length of the DHCP lease offered to the Orbi by the NBN Optical Network Terminator (ONT)? The typical lease is 86,400 seconds (one day). The typical pattern is for devices to attempt to renew the lease when 1/2 of the time remains (43,200 seconds) and to keep attempting to renew until it expires. At this time, the device must cease using the IP address because the IP address is invalid. If there is an issue with the DHCP process, one would expect the ONT to router connection to fail in almost exactly 24 hours.
This information is displayed in the Orbi web interface on the Advanced Tab home page by clicking on the blue box Connection Status.
Troubleshooting this sort of issue is not trivial, and may not be worth the time, effort, and inconvenience.
My experiment was as follows:
- Purchase a gigabit Ethernet switch that has the capability to mirror traffic from one port to another.
- Place this switch between the NBN ONT and the RBR750 router.
- Connect port 1 of the switch to the ONT, port 2 of the switch to the RBR750, and mirror port 1 to another port.
- Connect this port to a computer which has an Ethernet port that can run a data collection program. I purchased a gigabit Ethernet to USB adapter and plugged it into my Windows PC.
- Run a program on that computer to monitor the connection. (I used Wireshark - free for Windows, MacOS, & Linux).
- Define a Capture Filter on Wireshark to capture only DHCP and DHCPv6 traffic. (no point in collecting gigabytes of user data).
- Run the program for several days (I ran for about 10 days) or until the NBN to Orbi connection fails.
- Examine the resulting file to display the DHCP conversations between the ISP device and the Orbi router.
What I discovered with my Spectrum connection was this:
- Spectrum assigned the Orbi router a regular DHCPv4 lease of 86,400 seconds as expected.
After 12 hours, the Orbi asked to renew the lease, and the Spectrum modem granted a new lease for 86,400 seconds. This went on day after day. - Spectrum assigned the Orbi router a DHCPv6 lease of one week 604,800 seconds (one week). After 3 1/2 days, the Orbi router asked to renew the lease, and Spectrum again granted the same lease.
By this point, it is probably obvious that investigating this situation is not trivial. It requires additional equipment, effort, and the inconvenience of changing firmware back to version 7 until the connection fails.
If the DHCP renew process is the root of this problem, it should be relatively obvious. However, it is not clear (to me) how "having proof" would improve your situation. So what? NBN and Netgear "blame the other guy" and neither of them does anything about it.
Or, the explanation could be a miserable failure. The DHCP lease could be renewed as expected and there is some other reason that the connection drops once a day.
I do this sort of thing because I enjoy it. Most users ........ probably not so much.
Thank you for your response. This is a lot of effort that you describe here. But maybe it is worth it.
The lease is valid for 10 min. I was surprised about this. But even if I release and renew it shows me 10 min. I am with Superloop and it seems to be real and I am not alone with this. E.g. https://forums.whirlpool.net.au/archive/3xk0vwy9
With buying a gigabit switch, I am not so happy and wiresharking the connection is much effort.
As this dropout is every day, mostly overnight I would not relate this to the lease time of my IP address.
I also should mention that in case of the issue happen, the NBN modem needs to be restarted. I leave the ORBI untouched. The NBN modem is connected over UNI-D 1 to my ORBI.
Looking at the LEDs of the modem (https://apcoms.com.au/nbn-fttp-connection-box-status-lights/) the UNI-D 1 LED is flashing orange, which indicates 1GB. If the issue is happening, this LED is kind of pulsing orange and the alarm is showing red.
Then I power cycle the modem and a minute later all is back to normal again.
As this issue firmware depended, I would assume there is something wrong – or maybe Australian NBN is ‘special’. I have Fiber to the node.
To be fair, if NBN is stating the router is the issue, they were right. But Netgear does not give you support for longer than 3 months. They only offer you to purchase support after 3 months and there is nobody to talk to.
I also disabled IPv6 in advanced settings - Internet Connection Type, but this should not be related.
I do IT for living, so I enjoy watching Netflix afterwards and not troubleshooting my own stuff (too much) 😊 Spending much $$$ on a good ORBI system should have a more reliable outcome.
Something to try, install a non managed Gb LAN switch in between the NBN ONT and the RBR. Don't connect anything else to this switch, just the RBR and ONT. Load up v.21 FW and test. Something like a NG GS105 or GS305 series.
CAT5E or CAT6? There no CAT6E cabling.
I would agree, possible something specific with this ISP service, there ONT and v7 FW that is problematic. I passed this on to NG for review. No idea if or when anything will be done. There are known issues with v.31 otherwise so we don't recommend using that version.
Otherwise, keep v4 loaded and find something different at some later point or check on a FW update if and when that appears.
Good Luck.
HaJoS wrote:
You’re right, the RBR750 is a router…
The Modem is from NBN and belongs to NBN, this is part of the infrastructure here. The model is G-240G-P and details can be seen here: https://apcoms.com.au/nbn-fttp-connection-box-status-lights/
From this NBN modem, I have a CAT6E cable in the walls to the centre of my house which is about 15m long. I have replaced this cable 2 times with 2 new temporary cables, but this made no difference.
My ISP supports DHCP WAN, it does not require a login and it is not PPPoE.
I did a full factory reset after v 7.2.6.21 and configured it all from scratch.
My ISP did some investigation and they said my router is causing this issue, I should try with another router. But after rolling it back to v 4.6.14.3 it just keeps working fine for months without rebooting.
With any later FW version I get this issue daily, so the only logical explanation is that the firmware causes this.
Technically you are right, there is no CAT6E standard. However, It is a CAT6 cable which is about 10m long. As this was made for a data center it is kind of or supposed to be better than CAT6.
However, I was replacing the cable twice with a normal CAT5E cable and there was no difference.
I ordered a GS305 which should come today. I have also a managed switch here, a GS305E
I let you know how this goes.
