Newfoundland

The reason to check is you need to verify the results. A single sample test should always be checked ie. multiple data points to insure you have correct results.

Could you either run a ping cmd on both routers to compare the estimated base ping or use PingPlotter at the same target on both routers and take screen shots. with game servers it’s not so easy due to routing and it’s better to target for example Google and see what your base pin is like. You need to have no traffic when you test so we can see the results

How have you checked it other than another router?

The Asus would need to be set as AP mode and plugged into a lan port. Your ONT would go into the WAN port of the R3. You would then disable WiFi on the R3 Hi, try using PingPlotter instead of waveform. I’m not a fan of that site to be honest. PingPlotter is a good tool and you can test your connection with that with different traffic types too. I noticed you have a fairly good package, bufferbloat really only comes into it if your connection becomes saturated. Try PingPlotter wired up from either a pc or Mac and try first on default settings. You can leave it running and just monitor your connection to see how it is. Default is a simple ping test which will help to show you some results. the boot up time varies across the board. My Unifi stuff takes a while to boot up as it also runs test as well as preparing the network. It’s nothing to worry about as it’s rare you need to reboot. In fact Unifi takes ages to fully boot as it has procedures in the background but it’s not an issue. oh Unifi has plenty of bugs, more than Duma. It’s no different than most in the domestic market. It’s no way near the big boys when it comes to it but it’s cheap for what it is as you don’t have licences or 24/7 support. Ping in its self is a simple test of connectivity. It’s not a true representation of traffic but is handy for monitoring connection. Many companies use it as a way of logging connectivity. Have a play with PingPlotter, see how you get on but most importantly enjoy some games in the festive period.

Is that the client telling you this or the routers software? 20 mtrs is 65.6ft so 55dbi is pretty much impossible to achieve. Is this 2.4 or 5Ghz?

At range forget about speed it’s the retransmission that is your enemy. You should always try to keep a client under 70dbi as that’s the limit. many clients would still show a good connection at that rate but would be having errors with network collisions, the error of receipt and creating a higher latency across all WiFi clients.

I know what you are saying but I don’t get why you require that info. How important is a speedtest to you for example and what do you hope to achieve. We don’t know what subsets are running with the Duma router, for example OFDMA up and down, BSS Coloring, TwT, PMF. Eero will have these subsets.

What is the latency of that connection from client to wan ? Remember it’s only showing 1 way, there is no method on speed checks to check the communication standard protocols. Speedtests are basically pointless when you want to check how accurate the standards are followed

Both use official drivers and good manufacturing. A wifi6 router does not guarantee better throughput , WiFi 6 was all about connectivity. WiFi 7 adds more throughput and better connectivity. However not every WiFi 6 router supports or even runs some of the subsets.

It’s because we all have different environments and clients. WiFi is a strange beast, it’s always changing. As long as you have good return communication then that’s half of the battle won.

Most clients are 2x2 so you will struggle to get 900 over WiFi at a stable rate. An iperf test would be more accurate to show you the average you get because in reality traffic to and from client can go up and down. A speed test might show 800 but an iperf test might show you hit 800 as the highest point, 612 as the lowest and an average over a set time period of 690. In reality though these tests will change every time. clients vary too as does background interference. Hitting a gig on a 2x2 client at 80Mhz is never going to happen in a month of Sundays. what you want is a stable connection with no errors or retransmissions. Speed is not a way to test WiFi. You could get 800 with 80Mhz and 400 on a 40Mhz and the 40Mhz could end up more stable as there is less interference, slightly better coverage. Remember errors in connectivity slow down the network so you want a stable connection. the higher the Hz in 5Ghz the more interference as you are using more channels. This also affects coverage. you don’t want the router to ask to repeat the connection as other clients have to wait in turn. This increases latency within the network. Too close or too far away, too much transmit power all cause issues, WiFi has limitations, WiFi 6 was a small step to help but limitations will always be there. the secret is to make sure you have adequate coverage and that your clients are always within a good coverage zone. Yes clients determine connections and yes you can adjust connectivity with some APs but in general you want to map out your property and try to keep coverage equal in the hope you don’t suffer errors to keep the WiFi working at optimal. Remember a clients tends to not have the same connectivity range of a router so while the client sees the router it may well struggle to communicate back at range without errors. Speed tests are never going to show how great your connectivity is, it’s not a good way of testing in the long run. It’s just part of the picture. Try not to get hung up on these speed tests. Think about how stable you want your WiFi network, the clients and coverage.

Clients should be able to see each other, unfortunately I don’t have an R3 to test this out but they really should be able to communicate with each other without any need of a work around.

Have you separated 2.4 and 5Ghz and given them individual names?

They should be able to see each other, sounds like a bug or internal setting error.

You need to separate the networks like a guest network that should isolate clients from all other clients. so for example I have 4 networks and 2 isolated networks, one for guests on a voucher time setting and another for some equipment attached. It’s however not a Duma router. if this feature is not available perhaps worth thinking of adding it as a feature at some point.

To understand a little more why DMZ is a security risk think about your whole network which can be accessed through a client that’s compromised. It’s like leaving your front door open. it means in basic terms an intruder can access all your network from one client being a weak link and it’s really easy to do. always protect yourself.

Netgear use official drivers and will have better hardware, they are a big company so can manufacture and pay licences. throughput is not indication of performance, while you have a large property the secret to understanding WiFi coverage is that a router will always out perform a client. A client my well see the WiFi at range but you end up with retransmission errors where the client repeats its self which in turn slows down the network as other clients have to wait. You want good coverage, clients need to be within a good dBm to avoid errors. Now on hardware that offers say 8x8 or more as new spec increases this errors can be compensated more due to more connectivity of clients but on a 4x4 device you will get 2 clients at a time as most are 2x2. You could get 850mbps or more but it means nothing if clients have poor connectivity. it needs transmit power and you may well be better suited having correct AP deployment.

Run PingPlotter a bd see what happens and report the findings back here

That would be impossible to answer as everyone’s environment is different. You would need to map out your property to see what is required taking into consideration that a router can far exceed a client for communication so you want good coverage to avoid errors in transmit which slow down the network.

It should not be resetting itself or having to reboot because of a ram issue. I would guess it’s either a hardware fault corrupted firmware, or possibly a spike with its power supply. it’s a real shame that Netgear aren’t able to hang on to their logs after a reboot as that may help to indicate what is going on. really strange question, but is your MTU set at the correct value?

DNSSEC can be used on a router or home network but it will need some additional setup and configuration. The router will need to support DNSSEC and you would need to obtain and configure a DNSSEC-enabled DNS server. Additionally, your domain name would need to be signed with DNSSEC. Some DNS services providers like Google Public DNS and Cloudflare offer DNSSEC support out of the box. unless you have requirements on a server and as primary use of these routers are for gaming more than more complex home networking I would think using providers like Google or cloudflare would be better. Supporting it on a router menu is one thing but it’s meant for other purposes as above. https://www.snbforums.com/threads/any-reason-to-use-dnssec-with-quad9.82748/page-2 I’ve linked a thread for some info to help. Like many options like DoH is not as straightforward as it seems. The end user must make the choice which suits them. DoH for example may interfere with internal security protocols like IPS/IDS as it’s encrypted so can’t be examined.

Having 2 R3 routers would create a double Nat situation so may impact some services connected to the second router. While there are work arounds it’s not an ideal situation and I would wire up an AP to avoid congestion depending on number and types of clients and their use.

Hi, sorry to hear you had this issue. It really should not be like that. I’m surprised it’s not fixed as it’s a serious flaw. No router should revert to an open network. The only work around is a power back up that cuts in on a power outage but again it’s not ideal and the consumer should not be liable to pay for items to work around flaws. what routers have you looked at as a replacement.

Depends on network and hardware. Most routers nowadays have enough grunt to handle that throughput fine without QoS but older hardware will show its age. Ideally you want as little as possible regarding application processes run by a router unless you have some of the better architectures that can handle extra security, various QoS shaping, complex DPI and logging.

If your phone does not disconnect on a reboot then there’s a bug and it’s best to either do a factory reset using the rear pinhole method and set up as new or if that fails manually install the latest firmware again and do a reset afterwards.