RT-AC86U Defaulting to 100mbps on PC

[neonJZA]

Hey guys. For some reason the router keeps defaulting to 100mbps on my desktop PC. Here's what I know so far...
1) Not a cable issue, cable works fine on other devices
2) Not an issue with my PC, as when it's plugged into my Cisco switch it gets 1gbps just fine

I don't see any way to force 1 gbps full duplex, I tried https://www.snbforums.com/threads/alpha-asuswrt-merlin-382-1-alpha.41380/page-3#post-350450 but the console responds with:

admin@RT-AC86U-93E0:/usr/sbin# ethctl eth1 media-type 1000FD
Auto-Negotiation enabled, with capabilities: 1GFD
Link is down.

Anyone else have this problem and have any idea how to solve it? Super frustrating! Just got this thing today!

 

[neonJZA]

Hey guys. For some reason the router keeps defaulting to 100mbps on my desktop PC. Here's what I know so far...
1) Not a cable issue, cable works fine on other devices
2) Not an issue with my PC, as when it's plugged into my Cisco switch it gets 1gbps just fine

I don't see any way to force 1 gbps full duplex, I tried https://www.snbforums.com/threads/alpha-asuswrt-merlin-382-1-alpha.41380/page-3#post-350450 but the console responds with:


Anyone else have this problem and have any idea how to solve it? Super frustrating! Just got this thing today!

Fixed it. I had to modify eth4. Here's the command I used.

ethctl eth4 media-type 1000FD

 

[dosborne]

Fixed it.

I would be concerned though that the hardware autodetect was unable to appropriately determine the speed. While it could be purely a quirk in the detection, it could actually be indicating an issue. Even though you say it isn't the cable, I'd be inclined to try a CAT7 cable to rule out the possibility. Certainly something that would be easy enough to test and there are many well known autodetect issues. Personally I hate having to force things that should work, and I certainly like to do what I can to figure out the issue.

cable works fine on other devices

While this reduces the possibility of a bad cable, it does not eliminate it. Some devices output stronger signals, offer better shielding, etc. So are not all equal any not a definitive diagnostic tool. Make the investment on a good quality cat 6 or cat 7 cable that you can use to really rule out cabling issues.

 

[follower]

Fixed it. I had to modify eth4. Here's the command I used.

ethctl eth4 media-type 1000FD

It's a defective cable.

 

[neonJZA]

It's a defective cable.

What I don't understand though is why it would give me 1gbps full duplex on the two other devices I tried it on, which where my AT&T modem and my Nighthawk X6S. If they both auto-detected 1gb, and the ac86u didn't, how would that indicate a problem with the cable? Also, if it was the cable, wouldn't that be visible when I run speed tests? I'm getting 1 ms w/ 945 mbps download speed.

 

[Grisu]

+1 try with different cable only for testing, CAT5e or CAT6 is fine, no need for CAT7.

 

[Greg72]

Different devices chipsets are more forgiving. Some will go to Half-Duplex. Just get another cable and go for it.

 

[Greg72]

+1 try with different cable only for testing, CAT5e or CAT6 is fine, no need for CAT7.

Exactly. Cat-5e will work and al, of the jumpers that you get these days is going to be Cat-5e. No need for anything else in a home or very small business.

 

[dosborne]

No need for anything else in a home or very small business.

Agreed not absolutely necessary, but if you are buying a cable, may as well spend a whole $6 and get one that has been officially tested and certified. I personally did this 2 weeks ago to give myself comfort in knowing that it had been speced out, whereas the box of 2 dozen cat 5, 5e, 6 cables that I had were essentially "unknown" quality after years of abuse. (Note I did say cat 6 or 7 in my post)

The last thing you would want is a "marginal" cable to be used as your ultimate test device.

@neonJZA it could be as simple as dirt or oxidation on the connector, a wire that has been bent a few too many times, or a dozen other simple factors. Every port has a slight power variation and under the right circumstances and combinations can result exactly as you describe. As replacing the cable is cheap and easy, it is definitely worth your consideration. Even though your override may resolve the issue on the surface, it would be good to know for sure so that there are no performance / speed issues or introduction of failures, and excessive retries, on the network in general. Obviously up to you to decide what to do.

 

[neonJZA]

Agreed not absolutely necessary, but if you are buying a cable, may as well spend a whole $6 and get one that has been officially tested and certified. I personally did this 2 weeks ago to give myself comfort in knowing that it had been speced out, whereas the box of 2 dozen cat 5, 5e, 6 cables that I had were essentially "unknown" quality after years of abuse. (Note I did say cat 6 or 7 in my post)

The last thing you would want is a "marginal" cable to be used as your ultimate test device.

@neonJZA it could be as simple as dirt or oxidation on the connector, a wire that has been bent a few too many times, or a dozen other simple factors. Every port has a slight power variation and under the right circumstances and combinations can result exactly as you describe. As replacing the cable is cheap and easy, it is definitely worth your consideration. Even though your override may resolve the issue on the surface, it would be good to know for sure so that there are no performance / speed issues or introduction of failures, and excessive retries, on the network in general. Obviously up to you to decide what to do.

I went ahead and grabbed a new cable anyway since they're so cheap. I haven't had a chance to swap it out yet though as I've been pretty busy. Hoping it was the cable.