Ports 1-4 are standard Gigabit Ethernet ports that support 10/100/1000/2500 Mbps connections.
The Link LED on the left indicates a 10/100 Mbps connection when amber and blue indicates a 1 Gbps connection. For 2.5 Gbps connections, the LED on the right will illuminate white. If neither port LED is illuminated, then the connection is down.
Ports 2 and 3 support 802.3at PoE+ with up to 30W per port and a PoE budget of 40 Watts.
The PoE LEDs are located below ports 2 & 3 with the icon next to them. They will illuminate amber when a device connected to the port is being powered via Ethernet.
The SFP+ Ports support fiber optic and Ethernet transceivers with 1 Gbps, 2.5 Gbps, 5 Gbps, or 10 Gbps connections.
The Link LED on the left will illuminate blue when there is a 1 Gbps, 2.5 Gbps, or 5 Gbps connection, it will illuminate white with a 10 Gbps connection.
The Activity LED on the right flashes blue when there is activity on a 1 Gbps, 2.5 Gbps, or 5 Gbps connection. It will flash white if there is 10 Gbps network activity.
So it seems their devices are running OpenWrt with a heavily modified GUI and proprietary management software.
I like that the Route10 has 2.5Gbit ports with POE, but at this price-point I think the BPi-R4 offers better performance and support (assuming the Route10 could be supported by stock OpenWrt).
Because that's what the MT7988A supports... I agree that it's best to use a switch connected to SFP+ on the R4. That gives you most flexibility for ports/features/price.
I managed to remove the oem webui and reinstalled luci. The aarch64_generic architecture worked perfectly fine. Here is the overview page.
Had to change the opkg configuration file to include the generic architecture and all as well as change the repository urls.
If developers want any more information from my hardware feel free to reach out!
I havent figured out how to make the changes persistent yet. But considering this isnt a recompile of openwrt and just getting root access and changing the webui to luci, all drivers and hwaccel should be untouched. Im just returning control back to the user rather than the cloud. Ill run some more tests and see what exactly we can modify. Ill look into making a github repo of just reverse engineering this board.
Heya, been a minute but I have finally been able to successfully build my own version of openwrt. There wasnt much documentation so it took a lot of just trial and error but Ive gotten to a point where I am able to run a FIT uImage over tftpb. HW acceleration seems to work flawlessly as there is no cpu usage when running a speedtest at 2.2Gbps.
Awesome, thank you for your work!
Would you mind posting the results of iperf3/Wireguard throughput tests, and posting them? I am interested to see what it can handle at this point.
Its not ready enough to do throughput tests, but I do know hwaccel is working and 2.5Gbps throughput didn't even touch the CPU usage. I'm still working on the device tree source files as there are quite a few differences that may need changing, its just a trial and error process until I get it to a point that is going to allow the hardware to perform at its maximum potential. Not to mention the source tree I am using is full of issues that I am resolving along the way.
I should also mention I have no way of doing a ota firmware update, have to use the uart terminal to run the OS
