GitLab

interfaces table for auto creating interfaces, in case I need to back
them out in a hurry.

a URL for those, and we should fall back to looking for the image locally.

interfaces.  Also add a couple of generic interface types (40G and 100G).

1. We now allow lans to be implemented by a path. We did not allow this
   before, cause some of the sanity checking code was a pain to
   implement for lans. Well, no more sanity checking, the user is
   responsible for doing things correctly (after all, they are doing
   experiments with their own switches!).

2. We now allow topologies with more then one switch to be wired
   together. The wires between switches are marked as "trunk" wires,
   which informs the configuration generation code in libosload_switch
   to create the trunks and do the little tagged/untagged magic that is
   required on procurve switches. The same information is used to mark
   the the logical wires between switches as trunks.

   Aside: this stuff needs some work; we have spanning tree on by
   default, which causes the trunks to not work correctly. When I turn
   that off, things start working. So need some help from others who now
   about spanning tree stuff.

3. Serious kludging in the Interface and Port libraries related to
   choice of primary keys in the wires table. In order to insert a
   logical wire (or interface) that represents a connection setup by the
   apcon, we have to overload the primary key since the node_id1 side of
   the logical wires is the same as the physical wire to the apcon. We
   have to have overload the node_id2 side too, but that is really just
   a problem when wiring two switches together. Anyway, the kludge just
   maps card1 to a different id, and the Port library unmaps it. It will
   do for now, but really need logical wires to be done better then
   this.

1. The Geni path already supported this, just needed to add it to the
   web UI and plumb it through.

2. This is featurized, so on the Mothership only users with the feature
   will see this, not something we want mere users to do.

replace with new one that looks in the DB directly.

system announcements.

lookup.

1. Allow uuids to be used to specify reservations, change pretty much
   everything in the web interface to use uuid's so we stop exporting
   databases indexes to the client side.

2. Change RPC path to return a blob of data when approving a
   reservation. Ditto for initial creation, so that we can see precisely
   what the local cluster has done.

3. When a reservation is created/approved, insert an announcement in the
   announce system for that user, set to go off 24 hours ahead of
   reservation. Update that announcement when reservation is modified.

1. Implement most of issue #258: Add a start and end time for
   announcements. Also add a target uid for announcements.  Also add a
   purge option to fully remove announcements from the database, rather
   then retiring them. Add priorities for ordering announcements.
   Add update mode to modify an existing announcement.

   In other news, announcements now have a uuid and can be specified on
   the command line using their index or their uuid. For the web
   interface we now use uuids only instead of database indexes.

2. The web interface now polls for announcements so that users see new
   announcements without reloading. I noticed that a lot people stay
   parked on the status page and might not see them. Cleaned up DB
   queries and formatting to make is easier to use in an ajax call.

the new create image workflow.

to CreateImage().

We need to get rid of this!

Since BSD jail days we have attempted to partition up the UDP/TCP port
range among vnodes as jails and their host shared the same namespace.
Originally we supported a range of 256 per experiment which wound up
limiting the number of experiments we could instantiate. In order to get
a class up and running where we expected a large number of single-vnode
experiments, I reduced the range to 32 to allow more experiments, forgetting
that we pick a unique port per-vnode from that range to use for sshd.
So as a result I limited the number of vnodes per experiment to 32!

Did I mention that we need to eviscerate this mechanism with extreme
prejudice?

In checknode code for FreeBSD don't check the /dev/ad* device if it is a symlink.
[I think the a error in the test command for -c]

Closes #323.

with something other then their row index. Something unique in fact.

does not croak for them.