Q: Ethernet Ports Stop Working
-
Jun 6, 2016 12:49 PMby lawandreann,So far i've tried the solution posted in the first post on page 7 of the following thread, from user kickmyass: Ethernet has a self-assigned IP address and will not be able to connect...
i.e. deleting com.apple.alf.plist and rebooting the mac pro.
I'm not going to keep my hopes up but so far so good, ethernet port is fine, for almost two weeks, i think... Will post back if I see the issue again.
Other things i've done for the time being is connect the 2nd ethernet port and the wifi to the same network for redundancy. Network settings is configured to use Ethernet 1 as the primary network, and the router is routing outside traffic to Ethernet 2. In my situation this makes most of the traffic go through Ethernet 1 and so if Ethernet 1 starts breaking (self assigned ip symptom) at least my services for outside traffic still works, OS X will now use Ethernet 2 for all of its traffic and it gives me time to realize the problem is creeping up and reboot the machine before Ethernet 2 goes down. Wifi is there in the case both ports die really fast, since the machine is headless, so i can still remote in and reboot it as a last resort.
If deleting com.apple.alf.plist doesn't work, I'm going to give up. At least i have some sort of redundancy working on the network.
-
Jun 6, 2016 3:05 PMby wb5hvh,My instances have slowed down considerably. As I mentioned before, hooking up the 2nd ethernet port does not help my situation. I can turn on WiFi but that is not a fix for me - reboot remains the only fix. Not sure why the issue is happening less often unless one of the El Capitan updates has helped.
-
Jun 13, 2016 1:18 AMby MILJW002,I've been having the same experience for about a year. I've got a Mac Pro running El Capitan, but I think this was happening with previous OS X versions.
Testing so far.
If I try and copy a large amount of data (e.g. a 40GB VM) to my TC, this will trigger the behaviour within a minute or so.
If I connect to the TC using SMB (not AFP), then the connection appear more stable. The issue with SMB vs AFP is SMB enforces maximum file size limits.
If I reboot this restores the ethernet connection.
If I disable (make connection inactive, then active) this also restores the connection.
When the fault presents, the Mac thinks it's still connected, there is just no traffic flow. If I try and PING anything, no response.
Have also tried static IP addresses, using either ethernet port. Same behaviour.
It's like the TCP/IP stack plus the AFP client can't handle large file transfers.
-
Jun 13, 2016 4:59 AMby wb5hvh,A large transfer will sometimes trigger my problem but can't say that is the cause because it will occasionally happen with little activity. Also my ethernet activity can be a web download or inside my own network - there is nothing I can point to specifically. Reboot is my ONLY option but reboot always works. I normally use static IPs throughout my network but have tried DHCP with same results. The way I'm reading this, there must be several concurrent issues because of the various triggers and various solutions.
-
Jun 15, 2016 4:39 AMby Norton Chia,It's interesting as I only noticed in the past few weeks the ethernet ports on my Mac Pro (Early 2008) running El Capitan had self-assigned IP addresses. I only use one port, and switching to the other gave me the same result. I only noticed as I turned off Wi-Fi on the Mac Pro and then couldn't VNC remotely.
Just before coming here I found the com.apple.alf.plist trick and tried that. Not sure if it will cure it, but I note the last modified date was in 2011, to the day (but not hour)...
Anyway, so far it's doing what a reboot has done prior to this - given me back a reserved IP address. Will see how long it lasts.
