If we destroy the Blockstore DB state first and then the server-side object
destruction fails, we leave behind a dangling object and potentially Lease DB
state. But if we destroy the object first and the DB state removal fails,
we are left with a blockstore with no server-side object and thus we cannot
look it up in the future to retry the destruction.

We choose to go with the latter and just create a tiny server-side stub
object if the DB state removal fails.

This doesn't fit the real problem, but the failure is a lot less ugly!

Convert download buttons in manage_profile and show-profile to use the same save-to-disk method as genilib-editor.

Also fixes a bug where downloading the source for a profile always downloaded the first version rather than the currently-viewed version.

Hopefully, will make it easier for the numerous scripts that want to
mess with the swap device in /etc/fstab.

do reservations.

Also capture all ipmitool output for presenting to the caller.

Primarily I was after what was causing the occasional segfault.
That problem was caused by calling tmcc on a node that was free.
Seems we were derefing some NULL columns returned by mysql because
we assumed that there would always be a row in experiments for the
node in question.

Since I do need to call tmcd from the "pxewait" initramfs on Moonshot
ARM nodes, I cleaned up this assumption.

Two wrongs do make a right! Now that both bugs are fixed, let's see if
bootinfo will send the right events at the right times...

architecture, be sure to skip those when computing counts.

have to watch for that case. Not perfect but will do. Not sure why
we ever thought it was a good idea to allow type=class! :-)

and images.