BE30000 (AKA BQ16 Pro) not getting full bandwidth

[The Sandman]

I have a mesh set up using 3 ASUS BE30000 units. Originally I had 2 Gig internet (Frontier fiber) and had no problem getting full bandwidth at the main router (around 2300 MB (2.3 Gig) uploads and downloads).

Yesterday I upgraded my internet speed to 5 Gig. This was provisioned by an onsite tech and verified using an EERO unit (comes with the 5 Gig package) which showed 5.02 Gig download, 4.8 Gig upload). However, my Asus BE30000 won't go above about 3500 MB (3.5 Gig) download and 2000 MB (2 Gig) upload. Tried rebooting multiple times, changing some settings like changing DNS servers, AI protection off, Traffic Analyzer off, QoS off... raised me from 3.2 Gig to 3.5 Gig download. As far as I know I'm running factory settings. The EERO was brand new out of the box.

Any thoughts on why the Asus can't get the full 5 Gig while the EERO does? Hopefully it's just some setting or configuration I need to adjust?

Thanks!

 

[eibgrad]

I wish I had your problems.

Since you were unspecific about the EERO, I don't know if it's a fair, apples to apples comparison w/ the ASUS. It wasn't clear if you measured at the router itself (i.e., the router was in fact the client), or measured from the LAN. I only mention it because in the case of the former, there's no guarantee the router is as capable a client in this regard as a LAN client (it typically isn't). It's one of the common mistakes I see from users when they use the router's built-in bandwidth measurement tools; they assume it's an ideal client, when it often isn't. It's mighty convenient, but often misleading.

 

[The Sandman]

Thanks for the reply. The Eero is the Max 7… new WiFi 7 model with 10 Gig WAN port and speed was measured at the router using the Eero app.

For the ASUS, also used the 10 Gig WAN port and measured at the router using the ASUS app (Settings > QoS > Internet Speed).

Is there a better way to do the measurement?

 

[eibgrad]

Measure from a LAN device (ideally wired) through one router, then the other, and compare.

 

[The Sandman]

Sorry, could you be a little more specific?

My mesh is fully wired so there is an Ethernet backhaul. I have 2 PCs wired to the network at 2 different BE30000’s. Everything else is wireless, with only 1 WiFi 7 enabled device so far. You’re saying to check the internet speed at one of the PCs for example? Neither PC has 10 Gig ports yet although I intend to add them to one of them tomorrow.

Or did you mean something else?

 

[eibgrad]

Neither PC has 10 Gig ports yet although I intend to add them to one of them tomorrow.

Well then your performance testing is probably premature.

Your router's primary responsibility is to route packets between the LAN and WAN. Everything else (DHCP server, DNS server, firewall, VPNs, and any other *applications* (e.g., measurement tools)) are just a sideshow. In fact, in a business environment, you would NEVER run these services and apps on the router. Instead, you would run them on powerful servers, or in your own case, a powerful Windows or Linux machine, w/ NICs to match the LAN/WAN capabilities of your router, and ideally wired directly to that router. You then benchmark your router's performance using measurement tools on those machines. That's all that counts.

Instead, what users often do (at least by default and/or initially) is rely on measurement tools that are running on the router itself, if only because it's convenient, which can be misleading. Rarely is the router as efficient as any of those wired LAN devices. So you end up (effectively) measuring the efficiency of the router as an applications platform, rather than measuring its efficiency as a routing device. It's the latter that matters, NOT the former.

 

[The Sandman]

That all makes sense. It’s just irksome that the Eero sees the full 5 Gig bandwidth out of the box and the ASUS doesn’t.

 

[The Sandman]

I know I won’t see much or any real improvement with the 5 Gig versus the 2 Gig at this time. I’m planning on upgrading my older PC to a new one later this year so it will be 10 Gig capable… my newer PC only needs an add on PCI card. I also have a 10 Gig capable 8 bay Synology NAS. I won’t have a WiFi 7 phone until the next iPhone is released… ditto the iPad.. I do have a Samsung Galaxy Tab S9 Ultra that is WiFi 7. So my network upgrades are a work in progress.

The reason I upgraded to 5 Gig internet now is that they were running a very good limited time promotion. And I’m also not the most patient person around.

 

[The Sandman]

Update… installed a 10 G card to my downstairs PC. Main router is upstairs, which goes into a 10 G switch which feeds my 2 mesh nodes among other things.The downstairs PC is connected via the 10 G LAN port on its associated node. So all hard wired.

At first, it didn’t work well. Internet speeds on downstairs PC started at about 3.5 G down, 2 G up. Not 5 Gig but not too terrible. But as time went by each test became slower and slower until downloads were less than 100 M. During this time the speed on the main router went way down as well.

Tried rebooting everything, new driver for 10 G card, checking settings, etc. but nothing worked. Finally did a factory reset on the whole mesh system… took some time but afterwards everything worked much better, with download speeds 5.1 G, uploads 3.5 G on the downstairs PC. Still maxing at about 3.5 G download, 2 G upload at the main router. Upstairs PC only 1 G capable… getting 950 M uploads and downloads. WiFi 7 tablet seeing about 1800 M uploads and downloads.

So mostly good news… just not sure why the main BE30000 doesn’t register full internet speed or why upload is so much faster than download… 1.5 to 2 G difference.

 

[skypx]

Not sure if this will work, but did you enable jumbo frames on the node, switch and PC?