FOG Trunk boot loop when PXE booting from storage node.

[BardWood]

To be clear, this doesn’t happen if I’m pointing at default master. This is Trunk build 7102. My FOG portal tells me there are newer builds available but this was cloned from git about 2 hours ago. Both default master and the storage node are on the same subnet\VLAN. If I have DHCP hand out the IP of the default master it works great. Pointing to storage node IP results in Chainloading failed:

[george1421]

Just to be clear I think there is a programming error here.

But please check the storage node configuration. Ensure there an IP Address and a /fog path defined for Web root.

Also for clarity is the storage node a linux server or some other type of device (like a nas)?

[BardWood]

I didn’t customize anything when installing this but did add ‘/fog’ as my Web root so that should be right.

[BardWood]

Adding ‘(node IP)/fog’ didn’t make a difference.

[george1421]

@BardWood Is that for the storage node. The master node should already set.

[george1421]

@BardWood said in FOG Trunk boot loop when PXE booting from storage node.:

Adding ‘(node IP)/fog’ didn’t make a difference.

As I said this is probably a @Developers issue. But I wanted to make sure you had the web root defined for the storage node, because that is what is missing (outside of the IP address) in the url string.

edit just looking at the picture again, whats missing is the background image url. So that is why its failing. Can we assume the storage node is 172.16.40.101? /edit

[Tom Elliott]

@george1421 I don’t think it’s anything much.

Look at the http://service/ipxe/bg.png. Notice how it’s not got an IP address? My guess is 172.16.40.101 is the main server, and it’s trying to be told to look at the node for data. From what I can see, the only thing I can think of it being is DB settings for this node/server.

What OS is the main server running? What OS is the node running?

Assuming (I know I shouldn’t), that the main server IS the mysql server as well, it looks, to me, like it cannot setup the webroot information at all. It’s not connecting the to the Database, at least not properly.

This can be caused by mismatched user/pass pair for mysql user, or possibly it’s trying to authenticate using an incorrect password checking mechanism (as found in this post: https://forums.fogproject.org/topic/6828/fog-tftp-storge-node-trunk-6455/15#).

If I had to guess, though, I’d start simple. Just guessing you are running Ubuntu? If so, edit the /etc/mysql/my.cnf file and look for bind-address. Comment it by making that same line start with a #. Save and restart DB with: sudo mysql restart. I think all that’s happening is all requests are being blocked as they’re restricted to connect with anything OTHER than 127.0.0.1 (or whatever address is in there).

Again, these are all just guesses, so hopefully you can at least start to see what’s going on.

[Tom Elliott]

Mind you all the “editing/restarting” of mysql stuff needs to be on the “real” server, not the node.

[george1421]

@Tom-Elliott is there a way to do a qc check on the storage node during the install to ensure it can connect to the master node spl server. That would avoid this type of error?? Just as a suggestion?