Ookla's resident Speedtest

[akamaka]

i have no problem with 500/40, on AC86u.
View attachment 32752

But for test some time ago i got 1000/100 and there was no problem with that speed

I disabled all the scripts, rebooted and got UP: 414.97 / DW 20.00.
Prior to ttah the following scripts were running:
Skynet, scribe, nsrum, conmon, scMerlin, uiDivStats, uiScribe, YazDHCPm, Vnstat, Disk check script, Format disk, LED control.
any gues which one may be the culprit?

 

[Picard]

Hello,

We have network service of 300mbps and when testing the internet speed on the main router using the built-in ASUS speedtest, we never exceed 230 mbps. I also have the same result using the local ISP speedtest using wireless to the main router. I achieve the same result with the built-in ASUS speedtest when selecting the local ISP server as well as other servers.
Performing a speedtest with a wired client to the main router and using the local ISP routinely achieves 320mbps. Performing the same test wirelessly on the AiMesh node, we also achieve over 320mbps.
There seems to be a limitation in the main router that prevents the higher speeds both when using the internal speedtest and wirelessly using the local ISP speetest.
This is our configuration.

• RT-AC68u with latest Asuswrt firmware <- Main AIMesh Router. This router is newer and is hardware version C1 at 1000MHz
• RT-AC68u with latest Asuswrt firmware <- AIMesh Node wired by ethernet. This router is older and has no indication on the label for the hardware version.
• All settings are at default on both routers (I reset to factory and freshly loaded the firmware). I have all optional features off (QoS, AiProtection, etc.).

I would appreciate any advice to change any settings and achieve the same wireless speed on our main router as we experience on the node.
We may move this to a new post but I did notice that the speedtest in the firmware on the main router provides the exact same result as wireless on the main router. However, wired to the main router and using the ISP speedtest provides a higher speed. Therefore, there seems to be a relationship between the two tests.

Thanks.

 

[VANT]

@Picard
AC86u have dualcore 1,8GHz CPU. You have mistaken the routers

 

[Picard]

@Picard
AC86u have dualcore 1,8GHz CPU. You have mistaken the routers

Hello,
You are correct, both routers are AC68u.
I am using the latest beta software that addresses the security vulnerability: ASUS RT-AC68U Firmware version 9.0.0.4.386.41994

 

[deathnite]

i have no problem with 500/40, on AC86u.
But for test some time ago i got 1000/100 and there was no problem with that speed
View attachment 32752

What results were you getting on the router when you had 1000/100?

I have a 1000 Mbps connection, but with the integrated speedtest on AC86U I usually only get around 580 Mbps with no scripts running, the highest was 750 Mbps. While in the browser I get over 900 Mbps, same server.
It's probably limited by the CPU on this model, not powerful enough.

@faria has both AC86U and AX86U.
On AC86U he's got similar results to mine, while on AX86U he's getting over 900 Mbps.

those speeds are consistent with my rt-ac86u too. In the RT-AX86U Ookla speedtest reports 900+Mbps up and down.

 

[Picard]

Hello,
In our case we use two AC68U routers.
The key issue is that we experience faster wireless speeds using a mesh node than we do with the main router. The mesh node is connected with an Ethernet backhaul and it achieves the same speed using wireless or wired devices.
My goal is to achieve the same wireless speeds on the main router as we do on the mesh node. The observation that I have is that there seems to be a limitation on the main router because we have the same speed test results using wireless as we do using the internal ookla test. To be clear, a wired device connected to the main router and performing an internet speed test achieves the expected performance from the ISP.

Note that I performed the tests using the same device on both the main router and the mesh node to make sure I controlled the variables. Also, the wireless connection on the main router is capable of up to 500mbps with LAN file transfer tests and easily sustains over 400. There seems to be a WAN bottleneck in the wireless portion of main router for some reason. The common symptom is that the internal ookla test provides the same result as the wireless test and wired testing through the browser provides much higher speeds.

 

[drabisan]

Surprised to hear about slow on-router speed tests with the GT-AC5300, network is idled and trying different Ookla nodes?

Yeah, tried all possible options. Including only 1 client connected to my GT-AC5300 and no clients on node.
It's ok, I don't mind results are not accurate. I'm getting better results from laptop connected via wifi and nearly all times I'm reaching full 1G from wired clients. That matters the most for me.

 

[VANT]

What results were you getting on the router when you had 1000/100?

I have a 1000 Mbps connection, but with the integrated speedtest on AC86U I usually only get around 580 Mbps with no scripts running, the highest was 750 Mbps. While in the browser I get over 900 Mbps, same server.
It's probably limited by the CPU on this model, not powerful enough.

@faria has both AC86U and AX86U.
On AC86U he's got similar results to mine, while on AX86U he's getting over 900 Mbps.

cpu load under test is about 50-70% (550-700mbit/s) similar as when i use test from pc (900-950mbits), so i think that this is bug in this beta fw, not cpu limitation.

internal test

test on pc

 

[el pescador]

Where can i find the server id list?
The ones near me are run by tin pot companies and need to go a bit further to get a main isp server. Thanks