Apple Event: May 7th at 7 am PT
HTTP is now the default transport protocol for NetInstall in Mavericks Server.
When I use that I get a kernel panic on the target machine. Then it reboots endlessly.
If I override HTTP and use NFS (the old default protocol) NetRestore works fine.
My current target is a late 2012 Mac Mini.
Any ideas how I can fix this?
OS X Mavericks (10.9.1)
A guess would be that the HTTP service isn't configured/started properly.
Go into Server.app and make sure the HTTP service is running.
"Websites" claims to be running. Is that what you mean?
That's what I meant, yes.
That particular panic (di_root_image failed) usually means that something is wrong with the image. It's very strange that it would work under NFS, and fail under HTTP. Are you absolutely certain you booted the same image in both tests?
Yes I am sure. The target stops endless panic as soon as I flip the protocol from http to NSF. I only have one image. It works fine with NSF.
How can I confirm that http is actually working on the server?
I usually do it by doing something like this from a client...
curl -I http://bnesse6/NetBoot/NetBootSP0/NetBoot_DS_Test.nbi/NetInstall.dmg
or
curl http://bnesse6/NetBoot/NetBootSP0/NetBoot_DS_Test.nbi/NBImageInfo.plist
The first line should connect an pull header info, the second one should fetch the NBImageInfo.plist.
Be aware that this is harder when you have spaces in the pathname.
Apache was found to be crashing on my Mac Server.
Even though Server App claimed Websites was running it was not.
After a lot of horsing around the fix was found to be:
Stop Websites
Launch Terminal and enter the following command:
sudo serveradmin command web:command=restoreFactorySettings
Start Websites
Why does Mavericks NetRestore panic when using HTTP?
