This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
I bought a Orbi 970 system with two satellites in mid-February. Had it hooked up to a 5-Gig symmetric internet service.
Didn't have any issues for weeks as I only put a few devices on it while others sat connected to my older routers that were subnetting from the Orbi 970. As I slowly started moving devices over to the Orbi 970 network, random reboots started occuring.
I've followed all the thread here, Reddit, etc. and did everything to solve it without reaching out to support:
- Full reset twice.
- Bought a third satellite (so now router + 3).
- Wired all satellites to the router instead of using WiFi backhaul. (remove WiFi backhaul issues from the picture)
- Reconfigured WiFi channels used carefully to avoid overlaps with other routers I have + possible neighbor routers.
- Even had smart plugs turn off the router and satellites at 3 AM every day, and bring them back online in order (router first, and then satellites a few minutes later).
The reboots sometimes happened once an hour when the system seemed to be under what should be normal load. My wife was laughing at me constantly for the poor reliability of something that's set me back $3K+.
Applied all my engineering thinking to it, until one day, saw a thread where someone suggested that router overheating may be the cause and that they bought USB based fans from Amazon to put them over the router unit.
This sort of made sense because I did not have this problem when only a few (4-5) of my devices were configured to connect to it. First I put an Airthings sensor above the router to measure heat. Airthings showed external temperatures at 94.3 F at peak and an average of 91 F. The operating temperature for RBR maxes at 104 so it was plausible that the internals were coming close to that. The ambient temperature (room temperature) when all this was happening was between 68 and 72 F.
Fans arrive (all of $18), have them running in series on top of the router - boom, temperatures at Airthings right above the router drop to 74 F. I left the Airthings there - the highest I've seen is around 80 F. Everything's stable immediately - up and running for 5 days+ like perfection.
The real questions here:
- Do those of us who have this issue have defective units (poor thermal paste application on the processor)? I live in a colder area so I am sure people in hotter states are very likely to hit this more often.
- Is this a design issue? I.e. the fan on the unit is not powerful enough and pushing out heat. It's obvious a stronger fan is all that's needed.
- Do I care to ask for support and RMA? I don't know - I've the support subscription paid for.
My goal of posting this here is 2 things - (1) give people facing this a clear answer - empathetically excited to share this given my frustration for weeks prior to this, (2) get thoughts on what I should do?
- I could do an RMA and see what temps the unit puts out in the exact same setup without extra fans outside.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Netgear opened a support case on their own seeing this thread and reached out to me.
Gave them all the details they asked for to help them investigate.
Hopefully this leads to a solution that helps future adopters of this hardware have a seamless experience because the Orbi 970 is spectacularly good outside of this.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Netgear opened a support case on their own seeing this thread and reached out to me.
Gave them all the details they asked for to help them investigate.
Hopefully this leads to a solution that helps future adopters of this hardware have a seamless experience because the Orbi 970 is spectacularly good outside of this.
Support has all the logs they've asked for, for a couple of weeks now. The case is still open, so I guess engineering is investigating. Given multiple people are saying the same thing, I'm guessing some hardware revision is needed.
Not everyone who may buy these devices will have the same level of patience and troubleshooting drive to get to this mitigation so to keep the broader customer base for these high-priced devices happy, something needs to happen.
I'm not terribly bothered by the delay in getting back as I'm able to make it work with the fans, and the hardware is otherwise great. However, I got one of those new M4 iPads (which is another high-speed device), and I think it's randomly pushing the hardware over the edge even with the fans - my uptime is now 21 hours so at least one reboot has happened after a vast long stretch of no issues (when I was not home ironically).
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
I purchased the 970 bundle with 2 sats as soon as it was available (I think back in November of 23) and have had this same issue every since. I've gone round and round and round with support trying to solve the reboot issues. Finally noticed the unit seemed extremely hot. Put a fan on it, not a single issue for a month.
Recently I moved the entire network setup to an area with far more open space and thought it would be fine so didn't plug in the fan.... Nope, reboots and crashes 2 days in. Plugged the fan back in and all is working again. I feel this is a horrible design flaw. I live in Alaska and it's not a hot climate. 😂
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
I don't have any update from engineering, except support reaching out asking if this is still a problem, and/or if anything has changed. For how much they charge for these devices, it's very unfortunate that this wasn't discovered prior to starting large-scale manufacturing - adding fans could have been so relatively easy at design time and wouldn't have cost much in the bill of materials.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
This is the only known thread I have seen on this issue, and I also opened a case with Netgear on Tuesday over this issue. After an hour of analysis, they asked me to swap power supplies with one of the satellites and monitor uptime. So far after 2 days, no reboots, but mine have been pretty random over the last 2 months. I also measured thermal output of the RBE971, mine also shows about 90F under a relatively low network load. Have the RBE971 more out in the open now vice in an enclosure with a fan running over it, and it seems more stable right now. I also have applied engineering thinking, even asked support if I could try an engineering-level firmware with better logging to monitor the crashes (existing logging is just crap, only monitors minor events, which are lost when the router reboots), and they didn't sign up for that. I tend to agree there is an engineering flaw in the cooling system of the RBE971, and overheating seems to be causing the random reboots. But Netgear will not admit that at this point.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
While I started this thread about reboots, and I'd almost completely solved that with the fan sucking air up above the router, I had another issue with WiFi-6E where my 6E devices would lose connectivity with the devices losing even static IPs and get assigned 169.254.xxx.yyy IP addresses for the duration of internet loss. It would be completely random and only hit one device at a time usually.
A custom firmware did nothing to help with that (my newest iPad Pro M4 was replaced by Apple to rule out a Apple-facing hardware issue). Netgear decided that the logs from the router and satellites when this was happening was 'unusual' and initiated an RMA for the router.
I've filled in whatever was needed and I'm waiting for that process to go through.
For the rebooting issue, I don't think I have "reboots" but with the latest 'private' firmware, I do get random once-a-week internet drops even with the fan on top of the router. Affects everything including devices wired to the satellites. And the 3 satellites themselves are all wired to the router. The saving grace is that the recovery is very fast (relative to the original problem without fans) - 30 seconds or so. Measured temps when this happens are in the 74-degree range right above the router.
Because an RMA is happening anyway, I'll report if that does anything at all for reboots.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
FWIW, I have my RBE971 running in 'debug mode' right now after conversation with 2nd Level support, so that we can capture events in the debug log should it tip over this week, and send over to engineering.
How on earth did you get a 'private' firmware out of them?? 🙂 I'd love to help try that, and asked a couple of times, but nothing provided by Netgear.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
I think it was a known problem that they were testing a fix for, and since I reported the same thing, they probably thought it made sense if I tried out a candidate fix. I didn't ask for anything explicitly - I simply provide updates to them on behaviors, logs, etc. and follow through on whatever they ask in return. The open case on all of this is from February and ongoing - the reboot issue wasn't horrible since the fans worked, but the WiFi 6E issue is quite annoying as I don't have any mitigation for it.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Unintended CPU spike issue that further confirms (to me at least) that there is a thermal problem with my existing router (my RMA should arrive tomorrow) - sharing notes verbatim I sent to support to forward to engineering.
"I just wanted to share an update on the reboots problem with the existing router – even with the external fans, the router started rebooting several times a day.
I decided to turn on debug mode to help with logs, and something caught my eye in the router debug page. - The CPU load which was usually in the 5-10% range was reading 55%. - I refreshed the page frequently over 1-2 days and the reading was always in the 45% to 60% range.
What I think it was and what fixed it: - The problem started about the same time as I initiated the RMA. - The only coincidental thing I noticed was that Netgear Armor got deactivated on the router (assuming from your end, due to the RMA). - I have logs attached from one such reboot (on the router) attached - if it helps engineering find a previously unknown bug. - I reset the router fully today to see if it helped with the problem and it did! The CPU load has permanently dropped into the 10 to 15% range.
This unexpected high-CPU usage behavior did confirm one thing in my mind though - the reboots are caused by thermal events at the CPU. I could only measure 77 degrees F (max), continuously recorded, right above the router. I'm guessing high CPU usage spiked internal temperatures at/near the CPU much higher. After the CPU spike dropped, externally measured temperatures dropped to 71 degrees (about 6 degree temp difference outside between 15% CPU load and 60% CPU load)."
Additional comments beyond the notes above - I think I'd expect quality control on the routers to check artificially induced 100% CPU usage running in an max. designed ambient temperature environment for extended durations before shipping out from the factory. At least that's how I'd view this with an engineering hat on.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Your analysis makes sense.
I’m not sure how Armor works, but since your spikes stopped (apparently) after NG turned off your free subscription for the first router, it made me start thinking about the ticket I’ve had open with NG for 5-6 weeks related to 970 Slow Upload speeds.
I have noticed that I constantly get a rash of Armor “Vulnerability” Notices randomly. I get them when a new device joins (that’s expected, but then it appears Armor randomly rechecks all 60 of my attached devices (mostly IOT’s).
I know nothing about Armor, how it works or is supposed to work, but is it an NG “Cloud” based service”? If so, I wonder if the Armor tests originate from the Cloud and then some how upload result files to the NG cloud to either perform the test, or to “store the test results”? I never would have believed a 10GB rated router would have random CPU spikes up to 50-60%, but depending on how Armor works via the cloud, maybe those uploads (if there are any) might be very big files (possibly caused by Armor firmware issue?).
I have 5GB ATT fiber internet service (I’ve had the same Gateway for 3-years,) so I’m very familiar with the performance and symmetrical speeds (via the internal Ookla Speedtest, 99% of time up/down speeds in the 4800-5300Mb range, never below 3000Mb for the other 1% if tests). The ATT equipment I used for 3 years had only one 5Gb LAN router port, and both Satellites only had two 1GB LAN ports. So 6-weeks ago, I updated to the Orbi 970 with the 10GB ports on my 5GB service between my 3-year old Gateway (router in Passthrough, 2.4/5 WiFi bands disabled), but the Gateway has the ONT built into it, so I’m still using the Gateway unit for ONT and modem. Still showing same up/down speed test results.
After installing the NG 971-970, I immediately had upload speed issues 95% of the time. This same upload issue first surface on Community in mid-Feb.
100% of time, my Orbi download speeds are 4800-5100Mb. 95% of time upload speeds are 200-900Mb (over multiple hundreds of speed tests I’ve performed for myself, the Community, and NG Engineers). The other 5% (random occurrences, no config changes to router made, I’m still sitting in the same desk chair all tests are made from 6’ from the 971 router), I will get one upload at 4800-5100Mb. One minute before this test, and one minute later, next test will be back to 200-900Mb.
I’ve been sending NG de-bug files for a few weeks now. As luck would have, the debug file (and screen shots with time stamp) had one of the 5% upload speed jumps! I haven’t heard anything from NG since I sent them this info early last week.
With that said, I wonder if the constant 60% CPU usage causing your heat issues (my router is mounted on a NG wall mount 6’ above the floor, plenty of air circulation) could also be causing these slow Upload speed problems, depending on how Armor works. Maybe my “rare” 4800-5100 upload speeds some how catches an Armor “Cloud” testing break so upload speeds are “normal”??
What internet speed service to you have? It appears to me these slow upload speeds are somehow capped at 900Mb. If all the 971 routers function like this, if you have a 1Gb plan and see 900-1100 Download and 200-900 Upload (80% are in the 700-900 range) it doesn’t look as glaring as my 4800-5300 Down and 200-900 Up. Imagine when 10Gb speeds are introduced to residential customers. Those speed results will be jarring, and bad for NG sales!
Since NG already turned off Armor already, when you get your replacement, add some speed tests to your debug file investigation.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
To be clear, I've not received my RMA router yet. FedEx suggested I'd be getting it today, but it seems like it's not actually shipped yet (2-day shipping label was created July 1st, and no updates since then).
My existing router is running at 6-7% CPU load this morning just fine. Current uptime is 16.5 hours (so no reboots since reset; but with fans on).
Netgear Armor is not running anymore on my existing router. I have a paid subscription to Armor, and I think it got deactivated when the RMA was initiated (it was probably switched over to the router I haven't received yet).
When Netgear Armor was deactivated, something caused persistent CPU spikes in the existing router (probably a bug). Before this, it would be idling at the 5%-10% range and would still be stable with fans. With the CPU spike hanging at 45 to 60%, even the fans weren't helping.
In general, Netgear Armor does work because it identifies vulnerabilities in my Raspberry Pi from not installing updates on it for example.
Overall:
- Heat causes reboots (for sure on my router). Fans help with regular CPU load. - There may be a bug randomly causing CPU spikes (that Netgear engineering may be able to reproduce) - this is not related to the original cause of my reboots, but made it much worse and incapable of being managed with fans. Resetting the router seems to stop the bug from continuing. - Netgear Armor when it works correctly, may add some CPU load - not very significant usually. I don't know if the original reboots coincided with Netgear Armor doing something and spiking the CPU. - CPU spikes should not cause reboots up to the designed/documented support ambient temperature. - I have a WiFi 6E issue where devices get dropped for seemingly no reason - this is what finally triggered the RMA.
I have very fast internet - 5 Gigs symmetric up and down. 3 satellites all wired to the router. One satellite gets 10 Gig backhaul, the other two are limited to 2.5 Gig wired backhaul (originally wireless backhaul, but I wired it over MoCA to see if that reduced the reboots - it didn't). A desktop PC wired to one of the 2.5 Gig backhaul satellites gets 2.3 Gigs up and down on speedtest.
I have never had any internet slowdowns - either on connected devices or direct speedtests. 5.3 Gigs measured by the router. My iPad Pro (M4) sees 1.4 Gigs up and down over 6E.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Thanks for more info.
When I get some time this weekend, I’m going to try to disable Armor, (mine came with 1-year free, that’s why I thought yours was free also) and run more speed tests to see if there is any relation to Armor
I’m having the same 6E random disconnects. I can be sitting still with full 3 “puddles” if WiFi, and I can watch it instantaneously change to 5G cellular. Interesting thing, if I go into WIFI settings, it still shows the name of my Orbi Router?!?! Then, instantaneously, it’ll change back to full WIFI. Occasionally I’ll have to go back into the office and get close to router before it switches back to WIFI. I have several times had to power off/on my iPhone, and once I even had to reboot the mesh network.
I’ll be interested in any changes your new 971 makes. I ran into a similar issue when I was an “early adapter” to NG MR6550 mobile router. I had nothing but problems, on calls with BG constantly, finally after weeks they did suggest an RMA which I accepted. The replacement had different displays, config’s, etc. it was almost like they sold me a beta version first!! The replacement works flawlessly!! I’m hoping same for you.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
I've been having the exact same issues with router restarts on the 971. I've been in a wormhole of attempts over the last few weeks following all the suggestions and tries to fix it from forum suggestions and previously from NG support. This email chain is helping my sanity. I'm going to add a fan and disable Armor and see if it improves, I noticed yesterday during one of the reboots, that the router was hot. Thanks for the detail on your journey with this. You would think spending this kind of cash on a system would make the experience easier but this has easily been the worst equipment issue I've ever experienced.
Re: Orbi 970 Constants Reboots - Solved!!! (Router overheating is the cause)
Let us know if adding a fan on top and disabling Armor helps.
What Firmware version is currently loaded? What is the Mfr and model# of the Internet Service Providers modem/ONT the NG router is connected too? Be sure your using a good quality LAN cable between the modem and router. CAT6A STP is recommended. How many RBS are deployed? What is the size of your home? Sq Ft?
I've been having the exact same issues with router restarts on the 971. I've been in a wormhole of attempts over the last few weeks following all the suggestions and tries to fix it from forum suggestions and previously from NG support. This email chain is helping my sanity. I'm going to add a fan and disable Armor and see if it improves, I noticed yesterday during one of the reboots, that the router was hot. Thanks for the detail on your journey with this. You would think spending this kind of cash on a system would make the experience easier but this has easily been the worst equipment issue I've ever experienced.
So, I finally received my RMA units (looked like a brand-new box with router and satellites, everything sealed). Replaced everything and sent all of the original equipment including satellites back. Netgear Armor started over with a 1-year sub.
Great news overall - sharing the exact update I left in the support case --
(A) Overheating & connection drops. (good)
I can confirm that the new equipment does not overheat – I stress-tested it with a 52 GB game download (peaking at 250 Mbps down). This was one reproducible scenario that caused the previous router would have issues with.
Even though everything is fine, for hardware longevity, I still plan to run fans over the router.
(B) WiFi-6E issues (resolved through satellite movement) --
The second problem where my WiFi-6E devices in a particular room would randomly lose connectivity (that engineering thought the logs looked weird on).
This still happens on the new equipment. Same weird IP address assignment, etc.
But I resolved this myself in the end by moving a satellite and placing it directly inside the room with direct line-of-sight to all 6E devices (going from a wired backhaul to a wireless backhaul).
My devices near the satellite have higher pings (marginal at 10 ms) and marginally lower speeds (1.4 Gbps down to 1.2 Gbps up/down to my iPad) BUT the connection seems perfectly stable to multiple WiFi-6E devices in that room.
Re: Update to earlier post on overheating - RMA resolved the issue completely
Clarification to the post above - the peak download speed in the stress-test was 250 Megabytes per second (MBps), which would translate to 2+ Gigabits per second (2 Gbps).
