NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Ethernet speed: Router Mode vs. AP Mode
Hello, I have a question regarding the speed Ethernet connected devices in Router Mode (where my whole house switch is connected directly to my RBR850) vs. when my RBR850 is configured in AP Mode (an...
Would be helpful to know the speed of the ISP connection and the specs on the whole house switch. No satellites were mentioned.
As long as the discussion is confined to wired devices (no WiFi), then it is surprising that there would be a noticeable difference between these two approaches in terms of speed or responsiveness. It might be useful to quantify the results. There are tools for some platforms that can help "put numbers" on the impression.*
The big difference in the two approaches is that:
- Using AP mode disables all of the Orbi "router" features, such as Access Control, Armor, Parental Controls, selecting DNS servers, forwarding ports, using the VPN host feature, etc. In AP mode the ISP router features are what is available.
As far as the ISP router is concerned, the only WiFi devices it knows about would be those connected directly to the ISP router WiFi (if it is enabled). Everything that comes through the switch is (be definition) "wired". - Using router mode enables all Orbi router functions.
* For Windows, hrPING ( https://www.cfos.de/en/ping/ping.htm ) collects very accurate ICMP (ping) reports. As an example, I conducted a number of tests on my (older RBR50) network to contrast how 'wired' devices compared to WiFi devices. One test was between two HP desktop computers connected to the Orbi LAN ports. The results of 5,000 pings were:
| HP | |
| Router | |
| HP | |
| Max | 3.016 |
| Min | 0.243 |
| Average | 0.621 |
| 95% | 1.074 |
| 5% | 0.304 |
Some pings were very fast and some were (relatively) slow, but the average was about half a second and 95% were about one second. If the 'gut feeling' is that internet use 'feels different', one could run hrPING against some public web sites (as long as they respond to ICMP).
Another tool is iPerf3 (https://iperf.fr/iperf-download.php ) which is available for almost every platform (Windows, Mac, Linux, iPhone, iPad, Android). This tests the raw throughput capability of a link between two specific devices.
I sense that experimenting along these lines would be somewhat of a chore, what with changing the wiring around and changing modes on the router. It's the sort of thing most users would find not very much fun.
I sense that experimenting along these lines would be somewhat of a chore, what with changing the wiring around and changing modes on the router. It's the sort of thing most users would find not very much fun.
To say the LEAST...this process has been a MAJOR PITA and the work is still not done. I have learned...at least more than I knew before. I won't claim to know A LOT.
Would be helpful to know the speed of the ISP connection and the specs on the whole house switch. No satellites were mentioned.
I have Fiber service through CentruyLink, using their C4000XG modem/router. I have an RBK852 setup, our home is a ranch that is ~2300 sq. ft. on the main floor. The Modem/router is in the basement, but the RBR850 is an open stairwell between the floor and not behind any walls our under any floors. As the crow flies...the RBR is ~47feet from the RBS.
I have two switches hardwired to the RBR. One is a NETGEAR GS116 v2 16-port Gigabit switch which serves the whole house and a NETGEAR Nighthawk Pro GS810EMX. These switches have been used in both configurations but were wired directly to the ISP Modem/Router when in AP Mode.
As long as the discussion is confined to wired devices (no WiFi), then it is surprising that there would be a noticeable difference between these two approaches in terms of speed or responsiveness. It might be useful to quantify the results.
While the router was in AP Mode I averaged 909-710 down and 774-601 up for that period of months. Since changing to Transparent Bridged/Router Mode I am averaging 669-536 down and 265-188 up all over Ethernet. All results were calculated using Ookla Speedtest.
Most concerning from the results I am seeing over Ethernet is continued dropping of connected servers or drives and general sporadic web performance.