the SA about the slivers.

* More URN issues dealt with.

* Sliver registration and unregistraton (CM to SA).

* More V2 status stuff.

* Other fixes.

record to CMs when they asked for the Slice Authority (by prefix).
This will all get flushed when URNs are fully pushed out.

(including an update to make schemacheck happy), and client-side script
modifications.

Tested by me on a few different machines. Ok by Mike.

Added version tag to the top of schema files. Added a GNUmakefile in the 0.1 subdir so that it installs its files to lib/assign. Removed that functionality from rspec GNUmakefile and made it properly recursive. Changed xsdgen to automatically upload schema files to their proper web location on ops.

TODO: Change the subnode_of element to contain an attribute called parent_node instead of a text child node as it does currently.

comments (which are now out of date) if you only do it once.

daemon knows if the aggregate (slice) has been registered at its
Slice Authority.

the slice is colocated.

Made sure 0.1 files contain all of the additions we made in the runup to the GEC. Added request examples.

nsfiles table.

Got tired of looking in the DB.

project that already has admins.

add URNs to those ones.

Checkpoint of client at GEC. Hacks are in place for Georgia Tech (no CM there), Kentucky (temp), Wisconsin (temp), and CMU (temp). Supernodes are added when you drag a subnode onto the canvas. Other fixes.

own dhclient-script which ignores the exit value from dhclient-enter-hooks.
The result is that an Ubuntu node in an inner elab can wind up with a
1.1.1.1 nameserver if the DHCP replies come in in a particular order.
So we, um..."work around" that situation.

Also, ethtool is in a different location.

Make minimal attempt to detect a case where our nameserver="1.1.1.1"
hack might result in that value getting written to /etc/resolv.conf
and wreaking havoc with the rest of the boot.