Gui not loading after moving VM: Solved

Hey all,

 

I'm back up, figured I'd fill in my steps for those that run into this odd issue I stumbled into.

 

After removing my network.cfg file and rebooting, my system was somehow stuck in it's boot process. Once I got to my office, I did a hard reset and it came back up with a new MAC and IP. I went ahead and re-assigned it to it's previous IP with a new lease and began configuring my network setup to it's previous settings. As soon as I turned on VLAN's, all connectivity stopped, eth0 went down. I manually tried bringing it back up with:

 

ifconfig eth0 up

 

It wouldn't work. So I removed the network.cfg file again, rebooted, and started over. 

When everything came back up, instead of turning on VLANs first, I added a metric of 10 to the main eth0 interface. Again, everything went down, and I had to start over.

 

After the next reboot, I was able to get back up with my custom config with the below steps:

 

1. Instead of adding the metric for the main eth0 interface, I scrolled to the bottom of the page and added it as a route with my desired metric. Hit Done to save

2. Turned VLAN'ing on, crossed fingers. Saved. 
Everything came back up in about 30 seconds

3. Added my 3 VLANs (1 is for WAN, 1 is for a virtual LAN, 1 is for an isolated LAB LAN environment)
Crossed fingers, saved. Everything came back up!

 

Here was the tricky part. I like having my Unraid directly on the internet, but for this, I need my Unraid and dockers to utilize a separate internal subnet and VLAN'd WAN. We have multiple WAN addresses available to us at my office, so I took advantage and created a VLAN'd bridge to our WAN subnet so I could have my own external IP running through an internal virtualized firewall... The key was setting up the routing with a metric that when not functioning would direct network failover to the bare metal eth0 interface. In the past I think I got this working by accident. For whatever reason, if you add the metric to the interface directly, failover does not work. But if you add a routing rule at the bottom of the network apge with a higher metric (lower priority number) than eth0, it works. I verified this by running 

 

curl ifconfig.co

 

In multiple dockers after my virtual firewall spun on. All dockers reported the static WAN inside of my virtual network. Things are now back to the way they were. I still have no clue what brought everything down.

Changed Status to Solved

Changed Priority to Urgent