isadmin checks).

different queue sizes *and* units (packets or bytes) on a link, they
could have been bit by this.  But I doubt it.

  tb-fix-interface $node $link "iface"

what we get back from the DB is a pid_idx (or we can get a pid, if we
want...), so change a couple of comparisons.

name's don't match from resolve_link() to find_best_link() - the latter
returns only one link of each type, so it's too late to find the right
interface by the time we get back out to resolve_link(). Also fixed
a built-in assmption about the 'direction' that physical links go in.
The old code (explicity) assumed that links to from node to switch - now
we flip around our comparison if the directions of the links do not
match.

This was a bugger, even though it didn't involve much code, since the
'flipping' involves a 'parity' issue - it's really easy to mess up the
parity in several places and get it to work under *some* circumstances.

As a side effect, added some more information to the plink strucutre.
This turned out to not be necessary for the solution I finally came
up with - it was part of an earlier, abandoned one. But, it's probably
good information to have around in the future.

is hardwired, pass that hardwired boss value onto the image; i.e., create
/etc/emulab/bossnode in the filesystem(s).

i.e., we want the exit code of the program run by system.

ala freebsd7/rc.freebsd.  We need this currently for the new remote nodes
where we have to load partition images and not a whole disk image (because
we download a copy of the image to the hard drive and we cannot have that
mounted and reload the entire disk).  Anyway, with a single partition image
the swap partition will not be "loaded", aka initialized, and we are left
without swap.  Now the rc.linux script takes care of that on first boot.

packages I needed to install.

Makefile is not ready yet.

evproxy, which uses set_failover and set_connection_retries. The event
library had these as noops, so that was an easy change. Also, add back
in the async add/remove subscription stuff, which was already
implemented in pubsub but not hooked up from the event library.

In the tmcd/plab directory I purged all mention of elvin and changed it
to "event server". I also renamed the runelvin script to runevents.

In events/proxy I cleaned up the makefile and added evproxyplab to the
targets list since it should now build okay (no longer needs to link
against elvin stuff). Renamed elvindtest program to eventping, and
otherwise purged lots of "elvin" tokens.

instead of paths. There is a script called from database-migrate that
converts paths to logfile (db) objects.