Netgear Orbi daisy-chain in practice

[J23987]

Hi, new here; hope this hasn't been discussed already but I did some searching and didn't see anything.

Just wondering how the Orbi (in my case an RBK43) daisy-chaining has been working in real life situations.

Is it as effective as, and equivalent in behavior to, "real" mesh solutions or is it a hack that halves (or worse) speeds from distant satellites?

Since the system wasn't initially designed to support real mesh, wasn't sure if their implementation was as good as something like an Eero, Velop, and so on -- or if it was just a misleading hack so they could get some more customers on board.

Their documentation is very high-level so figured I'd get better information here.

Thanks!

 

[mjc775]

This is not exactly the answer you’re looking for, but it should give you a pretty good idea. In summary you should see no speed degradation at the furtherest access point.

Last week I setup 2 Netgear X6S Mesh Extenders in daisy chain for a client that wanted to extended the signal from an existing Netgear Nighthawk router on the north side of their house to a guest house about 150’ away to the south.

I put one X6S in a window on the south side of the main house about 50’ from the router, the other X6S was in the window on the south side of the guest house about 150’ between the two so they could see each other.

Speed tests were almost the same (only about 10% lower) in the guest house as they were on a wired computer in the main house.

Sent from my iPhone using Tapatalk

 

[Klueless]

I am very intrigued by your comment but I'm confused. What exactly were the results of your speed tests?

 

[mjc775]

I am very intrigued by your comment but I'm confused. What exactly were the results of your speed tests?

Top screenshot was taken near main Netgear Nighthawk router before extenders were setup. Bottom was taken in guest house near last daisy chained Nighthawk mesh extender after setup. Speeds were virtually the same. I also did a speed test on a wired computer connected to the Nighthawk router (not pictured), and speeds were similar.

Sent from my iPhone using Tapatalk

 

[Klueless]

Internet Speed Tests. Starting to understand better. What down/up speeds are you paying for? My guess is you're paying for about 20 x 1?

I've got to cut short but my guess is you're seeing some degradation in the daisy chain but your (s)low speed Internet is the "pinch" point.

Said differently I have a router in a window of one building and a range extender in a window of 2nd building (line of site, the windows face each other). I had 15 x 1 Internet. Was very happy, all devices, both buildings, 15 x 1.

Upgraded Internet to 400 x 20. Devices in 2nd bldg. now get 70 x 20 (not 400 x 20 like my Ethernet devices in Bldg 1). Point is yes, there is probably some degradation but it's still faster than your Internet so who cares : -)

 

[mjc775]

They’re in a rural area likely near the end of a hybrid fiber internet cable. I think their plan is 25x5, but on a wired connection at the AT&T gateway they average 22x1.5

Yeah, there’s a degradation of about 10% 2 extenders away from the source, but that’s nothing near the 40% average degradation you get from a regular extender that doesn’t use a dedicated backhaul radio.

On 2.4Ghz wireless with it’s 20-40 MHz wide channels, you’ll never get near 100 Mbps. With the wider channels on 5 GHz, you may get pretty close to 400 Mbps.


Sent from my iPhone using Tapatalk

 

[Klueless]

Yeah, there’s a degradation of about 10% 2 extenders away from the source, but that’s nothing near the 40% average degradation you get from a regular extender that doesn’t use a dedicated backhaul radio.

Actually, I do use a "dedicated" backhaul on my range extender.

On 2.4Ghz wireless with it’s 20-40 MHz wide channels, you’ll never get near 100 Mbps. With the wider channels on 5 GHz, you may get pretty close to 400 Mbps.

And that better explains my experience ... using 2.4 for all my clients in Bldg. 2.

I've heard the 10% figure before. Have you ever tested that in your environment? I'm almost tempted to re-engineer my setup to get better performance in bldg. 2.

 

[mjc775]

...I've heard the 10% figure before. Have you ever tested that in your environment? I'm almost tempted to re-engineer my setup to get better performance in bldg. 2.

In my test shown above, there’s about a 10% difference between the Internet gateway and the furthest extender. I have not tested beyond that.

I definitely suggest running both Wi-Fi bands in building 2. Give them the same SSID so that clients can pick the better band on their own.




Sent from my iPhone using Tapatalk

 

[Klueless]

In my test shown above, there’s about a 10% difference between the Internet gateway and the furthest extender. I have not tested beyond that.

I definitely suggest running both Wi-Fi bands in building 2. Give them the same SSID so that clients can pick the better band on their own.

If I did that I wouldn't have a dedicated backhaul and I'd incur the 40% penalty you mentioned earlier. Maybe I'll just buy a Linksys range extender with "cross connect", no that won't do it either because if the client connects at 5GHz then the backhaul will be at 2.4 and I'm right back to where I started.

Guess I'll leave it the way it is. 70 Mbps is still better than the 15 we had and we don't do that much work in bldg. 2 anyway : -)

(Guess the bottom line is I'd need something with 3 radios like some of the mesh systems have, one 5GHz for the backhaul and the other two radios for clients.)

 

[Easy Rhino]

klueless, what are the two routers you're currently using as extenders and how far apart are they?

You're already dedicating one band on the remote site as the backhaul, which seems good. But maybe flip it and using the 2.4 channel as backhaul and 5 for clients would be faster.

2ghz with 2 streams and 40mhz bandwidth ("N300") should get you around 150mbps under good conditions at close range.

 

[Klueless]

@Easy Rhino ;

Thanks for your comments. The two buildings are about 100 to 150 feet apart. The router and AP are mounted in facing windows with line of sight. I recently upgraded the Asus N66U with an RT-AC86U. The other is a Netgear 6150 range extender (AC1200).

I put in the range extender because the devices in bldg. 2 had intermittent connectivity at best. Some had no connectivity. At best none of the devices could even keep up with the 15 Mbps Internet connection we had at the time.

"Logic" told me to use 2.4GHz for the backhaul but at that time I had 2.4GHz clients that would not have been able to connect at 5GHz. Surprisingly readings from my WiFi analyzer suggested that using 5GHz for my backhaul would be as good as 2.4.

Bldg. 2 is an old bldg. with plaster and lathe walls and solid core doors. Walking around the old bldg. with my WiFi analyzer suggested I might have a couple of dead spots in Bldg. 2 if I used 5GHz for client connects.

So that's why I did what I did. In hindsight I wish I would have tried both ways back then and tested everything with actual data transfers rather than just WiFi readings. But I didn't.

I was curious when the one poster claimed only a 10% loss with Orbi but I think he explained it. He's got a 5GHz backhaul and a different 5GHz for clients. I've only a 2.4GHz for clients and I share the 5GHz backhaul with some of the clients in Bldg. 1.

I briefly toyed with the idea of re-engineering it all but I'm kind of a putz and decided to be thankful that things are working as well as they are. Our Internet was only 15Mbps. We upgraded to 400Mpbs. Clients in Bldg. 1 see it. In Bldg. 2 we're "only" seeing 60 to 70 Mbps but it's still 4X what we used to see : -)

As an aside if I connect a PC in bldg. 2 directly to the Ethernet port on the 6150 I see about 150 Mbps.

The PCs in Bldg. 1 are old to ancient. One of them running N over 5GHz sees about 150 Mbps. The others running over Ethernet see about 250 Mbps and the one good PC we have sees about 440 Mbps over Ethernet.

As another aside when we were 15 Mbps Internet we used to see daily usage of 15 to 20 GB. Now that we've upgraded to 400 Mbps we're seeing less daily usage, about 10 to 15 GB per day. I'm left wondering if the overhead of contention for limited bandwidth could cause nearly 5GB in overhead (errors, lost packets, retransmits, etc.) per day?

Again, thanks for the suggestions.