I just upgraded my Ubiquiti RouterStation Pro from LEDE 17.01.4 to OpenWRT 18.06.0. Everything went smoothly except that eth0 is now only able to connect at 100/Full (previously it would connect at 1G/Full).
I'm not exactly sure what is causing the problem -- Nothing changed on the upstream switch (TP-Link T1600G-28PS 24 port Smart Gigabit PoE+ switch w/ 4x SFP ports).
I installed ethtool and here is what I'm seeing:
`ethtool eth0`
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Supported pause frame use: No
Supports auto-negotiation: Yes
Supported FEC modes: Not reported
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Advertised pause frame use: No
Advertised auto-negotiation: Yes
Advertised FEC modes: Not reported
Link partner advertised link modes: 100baseT/Half
Link partner advertised pause frame use: No
Link partner advertised auto-negotiation: No
Link partner advertised FEC modes: Not reported
Speed: 100Mb/s
Duplex: Half
Port: MII
PHYAD: 4
Transceiver: external
Auto-negotiation: on
Current message level: 0x000000ff (255)
drv probe link timer ifdown ifup rx_err tx_err
Link detected: yes
I've tried forcing using ethtool to renegotiate, also tried forcing 1000M/Full on the switch.
I also tried setting the speed via ethtool directly, but it fails:
`ethtool -s eth0 speed 1000`
Cannot advertise speed 1000
And ethtool -s eth0 speed 1000 duplex full kills the link until I power cycle the device.
Any ideas? Is this possibly a bug? Or is there something else I can try?
So the issue is solved and was not the fault of OpenWRT at all. But I'm more confused than when I started.
@mk24 - yes, the switch as reporting a 100Mbps connection
The above quote had me thinking that maybe something was wrong with a physical port or with my wiring, since the 'partner' is the switch and I know it can handle Gigabit with no issues (it is a lower-end enterprise grade smart switch, after all). So I started swapping out cables and I was able to get normal gigabit again. But when I put those cables mechanically into place (in my rack), it would drop down to 100Mbps again (wasn't checking if it was full/half each time, but I know sometimes I was getting 100/full). Maybe a bad physical cable making intermittent connection... so I try another one and get the same result (out of the rack = 1G, in the rack = 100M)... now I'm thinking maybe it is some strange electrical noise source, so I start powering down non-essential systems in the rack but the results don't change.
After rerouting the cable a few times, I did manage to get it to connect properly at 1G in the rack, but I really don't know what is going on. I'm 99% sure it is not mechanical (i.e. bad connectors, bad cables, bad hardware). I still think there may be some noise issue or something odd happening that will come back to haunt me later. But my guess is that whatever the cause, I would likely need a $5k+ cable tester (that does really advanced diagnostics and measurements) to find the problem.
Keep in mind, this is a small rack, and the total cable length from the switch to the RouterStation Pro is ~4 feet (2 cables -- a 1-foot patch cable from the switch to the patch panel, a 3-foot cable from the back of the patch panel to the device, and a keystone coupler in between; all of these parts are CAT6 rated Monoprice cables (pre-terminated) and couplers, and about a year old in very good physical condition, never been mechanically or electrically stressed, and even replacing these parts had no effect -- it seemed purely positional in nature)
So I remain confused, but the problem was not OpenWRT in the end.