Describe the issue:
Async1 added a procedure on how to configure (faster) gPXE-based provisioning for kickstart. The procedure there contained a couple of workarounds, and there is a fix in for 6.2 on the engineering side. So, the procedure should be updated.
Engineering patch: https://github.com/theforeman/community-templates/pull/221
Suggestions for improvement:
This page explains everything (it's been updated):
* Except instead downloading the file from internet, use the file that
ships with Satellite, /usr/share/ipxe/ipxe.lkrn. Note that iPXE is just a more recent (renamed) version of the gPXE project.
* When working on the documentation, make sure to not that not all
hardware is compatible with iPXE/gPXE! It's quite subset of PXE-capable
* A new profile will *not* be shipped in Satellite 6.2, so the copy operation is still necessary to create an iPXE profiel. BTW symlink will NOT work because of chroot and SELinux; make a copy and fix labels in your documentation.
Related bug: https://bugzilla.redhat.com/show_bug.cgi?id=1215144
Moving to NEW and the default assignee to be triaged as the schedule allows.
I am just reviewing some of the older bugs in our queue to see what we can do to make some progress on them.
After looking over the description, it looks like what we need to do here is adapt the information you linked to into our documentation to provide some content on each of the scenarios.
Just to confirm - of A, B, and C, are there any that would not be supported with Satellite 6, or can we document them all?
Assigning to Melanie for review.
Melanie - the "Configuring gPXE to Reduce Provisioning Times" section is now located in the Provisioning Guide.
It looks like we need to do a few things to update this section, as follows -
* Add the steps of the scenario described to the documentation
* Add the list of requirements to the documentation
* Add three sub-sections to cover each of the scenarios A, B, and C pending
a response to the above question in comment #3.
* Adapt the text so that it assumes the environment is Red Hat, and to fit
the style of our documentation.
For example, scenario A would be a two-step procedure along the lines of
the following -
Copy the iPXE firmware to the TFTP root directory:
# cp /usr/share/ipxe/ipxe.lkrn /var/lib/tftpboot/
Correct the file context:
# restorecon -RvF /var/lib/tftpboot/
And a 'Note' admonition box that tells users not to use symlinks. In our
documentation, we always assume that SELinux is running as well, so we
tend to just tell users to perform steps rather than leaving it as an
*** Bug 1308896 has been marked as a duplicate of this bug. ***
It's not illegal per se to mention it, but I do agree with Russell that we should remove it. By including it in the documentation we put some burden, official or not, on the support team to try to support it.
So let's go ahead and pull it. The information is available from another site and not hard to find.
I can confirm this content is now merged.