Wrong Network Interface

[imagingmaster21]

I recently updated FOG from 1.4.4 to 1.5.0. When I go to capture a image off a host it gives this error:
https://dl2.pushbulletusercontent.com/U0KLNaXqAMHQxsyMAsmlwyVtQpAzAglH/20180517_100332.jpg

It seems it is trying to start Eth0 network interface,

The FOG webui loads fine and DHCP is not giving any errors I just cannot capture any images.

Also as a side note we did add more storage and moved the images directory to another drive with a SIM link, not sure if that is realeted or not.

IF I got to deploy an image it does try to do Eth0 and fails and ask you to hit enter and it ends up deploying no problem. My issue is only with capturing.

Also if you trace DHCP when a machine connects via PXE boot it shows it by IP and says it does not accept TFTP options.

[imagingmaster21]

@wayne-workman

Going from 1.5.0 to 1.5.2 resolved the issue.

[Wayne Workman]

@imagingmaster21 The eth0 thing is fine - this is the local NIC on the host, not the fog server. Your problem is with DHCP. What serves DHCP in your environment?

[imagingmaster21]

@wayne-workman DHCP is dealt with on our FOG Server. I made sure DHCP settings were the same as form before we did an upgrade.

[imagingmaster21]

@Wayne-Workman
Here is what my DHCP looks like.
https://dl2.pushbulletusercontent.com/sqF0fk0cF5pcqG4a0ZesbQK5ooEeBAet/20180517_120927.jpg

https://dl2.pushbulletusercontent.com/sqF0fk0cF5pcqG4a0ZesbQK5ooEeBAet/20180517_120927.jpg

https://dl2.pushbulletusercontent.com/z8gfNItC5aXxVxn92EgdjkEX1BqVVQHX/20180517_120911.jpg

[Wayne Workman]

@imagingmaster21 Check to see if dhcp is running like this systemctl status dhcpd -l

[imagingmaster21]

@wayne-workman
https://dl2.pushbulletusercontent.com/J08p9FQbsYLXGwN5l8vqu3Z98q9LnLhP/image.png
Is the status.

[Wayne Workman]

@imagingmaster21 Read through this: https://forums.fogproject.org/topic/11870/failed-to-get-an-ip-via-dhcp-only-on-a-certain-switch

[imagingmaster21]

@wayne-workman This worked before we upgraded from 1.4.4 to 1.5.0. We haven’t changed anything on the switch and all the settings are correct on the switch I verified those before.

[imagingmaster21]

@wayne-workman
The switch is… Pvst+ It’s basically a unique spanning-tree per vlan, but its running standard spanning-tree over them (not rapid)

[imagingmaster21]

@wayne-workman
I just tried this on a switch that another FOG Server running the same version worked on. The FOG Server that I am having issues on still does not work on the known good switch. So it is not the switch.

Any idea on what the culprit could be in DHCP?

[imagingmaster21]

@Wayne-Workman @Sebastian-Roth
I was looking at this here:
https://forums.fogproject.org/topic/7256/no-dhcp-response-when-trying-to-inventory-a-host-fog-trunk

I seen about trying a standard switch to see if it worked. So I grabbed a 5 port switch and still did not work. Here are some more screenshots:

• https://dl2.pushbulletusercontent.com/TjpLuWB0KKYGzdG41NsrvEeg1JoKErgy/20180517_133337.jpg
  -https://dl2.pushbulletusercontent.com/t1T9aZUgVR6dnnp23c74YsfDMYaoH9zk/20180517_133341.jpg
  https://dl2.pushbulletusercontent.com/zmul5sOnbxVi1AsknychwMCzYWxezeuQ/20180517_133343.jpg

I did use trunk to do the upgrade fro 1.4.4 to 1.5.0.

[Wayne Workman]

@imagingmaster21 Are you sure you have 1.5.0 installed? The latest stable is 1.5.2 and if you’re not using that, I would recommend you try it.

[imagingmaster21]

@wayne-workman
I am running 1.5.0-RC-9, would upgrading 1.5.2 fix this potentially?

[imagingmaster21]

@wayne-workman
Upgraded the version to 1.5.2 and capturing and pulling the image seems to be working correctly now. What would have cause this issue, just a bug in 1.5.0 that was fixed in 1.5.2?

[imagingmaster21]

@wayne-workman

Going from 1.5.0 to 1.5.2 resolved the issue.