testbed boot process and the freebsd pxeboot/MFS kernel. I'm hoping
this will be useful enough for users to load their own disk images,
but we shall see. Usage is simple:

	/usr/testbed/bin/node_admin on|off pcXXX

The "on" state boots the MFS, and the "off" states sets it back to
normal.

mounts).

eid in the portmap table.

reboot.

VLANS problem and ensures that the node state is cleared by nfree.

we don't have a root user in the DB. Perhaps this is bad?

Also made the 'forbidden character' message somewhat more helpful.

Changed request behavior to try to fill the most full of incomplete buffers, rather than a random one.
Also changed MB request to be a bit cleverer (tries to find contiguous received blocks and grow around those)
Changes resulted in less "bimodal" behavior and lower overall average runtime.

(now requests from random bucket, doesnt give priority to first bucket)

to enable debugging output.

Make snmpit fail loudly when it can't get the vlan edit buffer. Usually this fails for one of three reasons: a) The community strings are wrong on the switch, b) some previous snmpit run left the buffer taken when it unexpectedly was killed, or c) someone else really is using the buffer.

VLANs it is to operate on. This is only a temporary multi-switch solution, as
it only picks one switch to operate on.

Also, added a sanity check in snmpit_cisco.pm. Now checks every port that
it controls, to make sure it's on the right switch. If this sanity check
fails, dies to prevent any damage to existing switch configuration.

that assign does not handle well.

re-writing and formatting changes to make them more maintainable.  Also
re-worked command line handling to use Getopt::Std.

Add the days idle (web and users) to the long listing of users.

Also moved most subroutines to the end of the file, to match the style of
our other scripts.

I added a DB table to record last Web login, and I've added a backed
program (lastlogin.c) to get the lastlogin information from
users:/var/log/lastlog (mounted on boss:/usr/testbed/usersvar). These
two bits of info are now displayed in the user infomation page.

if its "spin", it was Jay,
if its boring, tedious detail, it was Mike

one pid/eid to another. This avoids names with duplicate vnames, which
screws up named (via named_setup) cause we get multiple/conflicting
CNAMES in the zone file.

Some minor cleanup/commenting as well.

Also, a few more access checks. Only admins are able to run 'inuse' (to prevent
users from seeing each other's experiments.) You are also only allowed to run
tbreport for experiments in projects that you are trusted by.

	Usage: os_load [-s | -w] [-r] [-i <imageid>] <node> [node ...]
        Usage: sched_reload [-f | -p] [-r] [-i <imageid>] <node> [node ...]

The imageid is now an optional argument. After continually forgetting
what imageid to use, or just plain forgetting the argument, and having
it try to load imageid pc53 on pcXX, I decided this interface was
bogus. With now imageid, select the default imageid for each node
provided. This is actually convenient since you can load multiple
types of nodes in one shot.

Fixed case where virt_nodes has no osid because they didn't choose one, and asked for a class of nodes instead of a type. now if they don't have an os, we choose the default os for the type of node that was chosen.