NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
RBR750 Not Reconnecting To Internet
This week I was away on holidays and was hoping I could rely on my HomeAssistant based security system with motion sensors / cameras etc. Many of these rely on WiFi / Internet. However two days b...
FURRYe38 wrote:Does the front LED on the RBR turn PINK when the RBR doesn't seem to connect to the Internet?
Ok so the GS108Ev3 is a smart or semi managed switch. These have been known to cause problems on Orbi systems with RBS ethernet connected. A GS108 or GS308 with out any letter characters after the model # are non managed switches and would be recommended in between the RBR and ethernet connected RBS. One thing you can try and check on the 108Ev3 is log into it's web page and disable any IGMP and Green Ethernet protocols.
Something to try as well, find a GS105/108 or GS105/308 non managed switch to swap in place of the 108Ev3 switch.
I've been testing a 850 series here with ethernet connected RBS and have found some CPU creep in upwards of 30% with two RBS connected. However my testing was wiped out this last Saturday with a home power bump that caused the ISP modem and RBR and RBS to power cycle and reboot. I was waiting for the RBR CPU usage creep to get high enough to see what would happen. GS108 switches are in use here. Not ruling this out as a cause of your RBR not connecting, however could be on contribution. I found that with RBS wirelessly connected, didn't cause CPU creep or having 1 RBS ethernet connected didn't see CPU creep upwards, adding the 2nd one did cause the RBR CPU creep upwards after the Hop Count on both RBS went from value 1 to Blank. Unknown why this happened and the results causing this creep.
Something to try, turn OFF one of the RBS and leave the other connected via ethernet, after checking the two items on the 108Ev3 options.
If the RBR is going pink when you notice no internet, might also be a problem with the RBR and the ONT failing to continue to work together and possible something is happening on the ONT side that stops and the RBR can't regain the connection. Something to check the front LED and check the Show Stats on the RBR for the WAN section. If you see the uptime reset here and the uptime on the RBRs debug page doesn't show same change, then the WAN connection stopped working here or was reset. I might ask the ISP to check the ONT logs and there ONT service logs from you location the next time this happens to see if anything appears to them when this happens again.
Need to find out why the RBR CPU usage is so high. Should be down the low tens or teens.
Thanks for this info, its great. I'll first try to change those settings on the managed switches I have to see if it helps. Do you know if anyone at Netgear is actively looking to fix those issues if they are known ones / possible to reproduce? That'd obviously be the ideal solution and would mean I could rule those in/out if a firmware fix was able to resolve at least those known ones.
For me when I came home, both the RBR and RBS's were a solid purple type color. My worry is, if they were having some issue with connectivity to the ONT, even if it was the ONT fault, why wasn't there some internal restart of whatever module/service (I'm not sure how it works) inside the firmware that would have meant the hard reboot from my side was needed.
I guess without being able to see some permanent log that survives a reboot I'll have to try to get into the router logs before rebooting it next time, possibly via the LAN if eg my PC can get to its IP over the LAN.
Yeah that will be interesting to see why my RBR is so high in terms of cpu usage. Even as we speak its still at 37.2% and almost nothing (that I'm aware of) is happening in the house (I guess something is causing it that I'm not aware of so I'll start shutting things off/on to see can I figure out whats causing it to be so high.
Does any other debug type of page show any sort of breakdown of that percentage to give any hints as to whats causing it? Does the debug page show a very up to date view of the cpu %. eg if I shut down a device, would it in theory affect the debug.html cpu usage figure fairly quickly or is there any lag / delay.
One thing to note, I remember when searching a few months back to see if I could eliminate the use of my ISP provided modem, as to whether the Orbi RBR could support the VLAN tagging that my ISP uses. It seems the Google WiFi doesn't support it but the Orbi does. But I vaguely remember reading some threads that using this feature of the Orbi router does in some way slow it down.
Is this in any way related to its high cpu usage, untagging/tagging this vlan id that my ISP uses as traffic comes and goes? Not that it causes any issues for me on a daily basis of course, and I guess also shouldn't cause the complete outage that I had, but just might explain the high cpu usage?
So within the routers settings under "Advanced" -> "Enable VLAN / Bridge Setup" I have set "By VLAN tag group" with VLAN ID of 10 set so that I can connect to my ISP directly. Does this also show high CPU for you when used?
It's possible. My ISP doesn't use VLAN tagging as it's a Cable ISP. So it's all DCHP here.
Only way to check that would be to put the ISP router online, use the ISP router to configure any VLAN tagging, then use the DMZ on the ISP router for the RBR in router mode and then monitor CPU usage.
Possible that this VLAN tagging and the ISP service at the ONT with the RBR is not working right and causing higher CPU usage and eventually stops if the CPU usage gets too high.
- Yeah I reconnected my ISP router and set the RBR as DMZ. Now the debug page shows 7%. Wow 30% is a lot of cpu to spend removing vlan 10 from my ISP even when 'idle'. Should it be considered a bug that it's using so much cpu. It'd be great if there was a fix for that. I'm more than happy to provide logs/captures to help with any possible fix/improvement.
I'm still not sure of course if it's related to the original issue I hit but interesting. - Well this is annoying it's back to 37 percent. I could have been hit with chrome auto complete and went to the wrong debug page earlier. So even connected to the ISP set in DMZ mode to the RBR IP, sitting next to it with a Cat 6a it's using 37 percent. Of course also happy to help debug it in any way.
Ok so something else is causing high CPU usage.
Is Armor, Access Controls and Traffic Meter enabled?
What devices are all connected currently to the system?Still using the LAN switches in between the RBR and RBS?
I might turn OFF one RBS and maybe let one RBS wirelessly connect to the RBR if it will and see if this changes the CPU usage.
I might turn OFF any and all devices and go with a limited few devices and see if the CPU usage goes down or maintains. Might need a reboot to check this again. Not sure if the CPU usage will change or not being in this kind of state.
Possible something on the WLAN or LAN side or switches maybe a possible cause as well. If you can get the RBR and RBS down to limited use and limited devices connected, then if the CPU usage changes downwards and get to a low value, then graduate adding back in devices while monitoring the CPU usage for changes in direction. Might find a culprit.
And you said you factory reset the system at one point right?
Yeah, armor, access controls and traffic meter are disabled. Sure I'll try to actually disconnect absolutely everything, even the RBS's and as you say bring them in one by one and report back.
