Various nfstrace changes that have been sitting in my tree for awhile.

	* GNUmakefile.in: Do fs-install in the sensors subdir so the
	nfstracer gets installed.

	* sensors/and/and-emulab.priorities: Add some more daemon uid's to
	be excluded from auto-nicing.

	* sensors/and/and.c: Ignore invalid uids/gids in the config file
	instead of dying.

	* sensors/nfstrace/GNUmakefile: Makefile used to generate
	nfsdb-create.sql.

	* sensors/nfstrace/GNUmakefile.in: Some more installation stuff.

	* sensors/nfstrace/nfsdb-create.sql: SQL used to create the nfsdb
	database.

	* sensors/nfstrace/nfsdump2db: Bunch of bug fixes and cleanup.

	* sensors/nfstrace/nfsdump2db.8, sensors/nfstrace/nfstrace.7,
	sensors/nfstrace/nfstrace.proxy.8: Start at some man pages.

	* sensors/nfstrace/nfstrace.init.in: Try to detect the interface
	to listen on, not perfect though.  Add a restart handler that just
	restarts nfsdump2db.  Some other cleanup.

	* sensors/nfstrace/nfstrace.proxy: Some optimizations for
	resolving file names.

	* sensors/nfstrace/nfsdump2/*: Only print summaries of read/write
	packets and start a separate thread to read from the bpf socket.

	* tbsetup/tbswap.in: Stop transferring nfs accesses to boss' db
	until we figure out what we want to do with it.

logs.

Finish off changes to consult the os_boot_cmd table when setting
def_boot_cmd_line.

	* db/libdb.pm.in: Add TBGetOSBootCmd function that returns the
	boot command line for an osid/role.

	* sql/database-fill-supplemental.sql: Add some queries that
	initialize the os_boot_cmd table.

	* tbsetup/assign_wrapper.in: For nodes that need linkdelays or
	inner elab boss nodes, set their command line using
	TBGetOSBootCmd.  Since the TBGetOSBootCmd function needs a "real"
	OSID to work, we resolve the nextosid for any meta OSIDs.

	* tbsetup/elabinelab.in: Add os_boot_cmd to the list of full
	tables to be dumped.

	* tbsetup/ns2ir/elabinelab-withfsnode.ns,
	tbsetup/ns2ir/elabinelab.ns: Remove boot command lines,
	assign_wrapper does it instead.

Fix up the plav event proxy a bit.  It was sending RELOAD notifications
when it should have been scheduling UPDATEs.  The RELOAD event handler was
also updated to schedule and update.  Replaced some sprintfs with snprintfs,
and added a log file and some helpful log output.

pa-wrapper: tell libtmcc we don't want a cached "plabconfig" result.

monitor. (output version 2)

Wrap read(), recv(), and recvmsg() so that we can detect EOF on a
socket by a zero-length read.

Run a function to announce the closing of sockets at exit time, in
case any are still open.

Wrap accept() so that we can deal with traffic sent (NOT recieved) by
servers as well.

Fix minor bug where I was forgetting to ntohs() a port number.

First, treat the return from read() as an int instead of an ssize_t.
The latter is the official return type of read, but it's unsigned,
so when read() returns a -1 on error, it looks like a huge positive
number. Stupid. Whose fault is this? POSIX's?

Second, make sure we never try to read on the listen socket.

The result of these two bugs was that I was reading from the listen
socket, and it was failing and returning -1, but that looked like
a successful read.

Work in progress, just want to snapshot it before I do something stupid,
like accidentally delete it.

running as root, processes running as a regular user can connect to
it.

monitor.

on sockets, so that we can correctly handle multiple sockets between
two hosts.

The version of output used is controlled by the new
LIBNETMON_OUTPUTVERSION environment variable. The old output is
version 1, which is the default for now. The new version is 2.

Version 2 is still in ASCII. The address format has chaged from
'remoteip.remoteport' to 'localport:remoteip:remoteport'.

Version 2 will soon get the ability include information about when
sockets are created and destroyed.

table.

the socket buffer to be that size (in bytes) by setting it ourselves,
and not letting the caller change it.

Added trunk stats lookup (given a member port) to and added snmp-if-deref.sh
to the CVS repo.

Bump rootball version - event system now supported

Fixes for running the program-agent properly under plab.  It works!

so they can be viewed with xplot.  Loghole does this by merging the
capture files for the "pre-shaped" sides of a particular link and then
runs that new file through tcptrace to generate a bunch of graphs.
The merged capture file and the .xpl's are dropped in the experiment's
log directory under separate directories for each lan/link.

Folded elvind, plab-adapted evproxy, and program-agent into plab startup.

rc.progagent: Do run on plab, and do so via the wrapper.

GNUmakefile.in: Add new plab startup scripts to rootball

rc.inplab: hook in the elvin startup script.

pa-wrapper: Sachin's program-agent wrapper for running on plab.  Tries to
            keep the agent connected to the local elvind.

runelvin: Sachin's script to fire up elvin and evproxy on Emulab.  Makes
          sure elvin starts up on an unused port if 2917 happens to be
          in use.

Add svcslice bit to plabconfig

for pelab use.

reporting does not start until the first quanta in which  traffic is
recieved, and exits when the number of bytes in a quanta goes back
down to zero.

Also change quanta to 500ms, as suggested by Jon.

mode, or the db entry gets marked as invalid.

usage of the batch daemon on swapin.

usage of the batch daemon on swapin.

an alpha prefix and a numeric suffix which sorts by the number.

Changed monitor to accept an optional argument specifying the ip-address or name of the stub. If this optional argument is not provided, the monitor looks it up in the mapping. Changed run-monitor-libnetmon.sh to provide this extra argument. Note that run-monitor.sh is now deprecated. Always use run-monitor-libnetmon.sh

Added a new command line options '-r' to represent
running on planetlab:

- If -r, it sets the connection_retry to 3, sets
  idle period to 60, and don't daemonize it.

Added 2 new programs:

1- elvindtest: a simple program to verify if a particular
    elvindserver is up or not

2- evproxyplab.cc: evproxy for planetlab nodes. It subscribes
   to the following events:
   host = node's ip addr
   objtype = EVPROXY
   objname = __$nodeid_proxy
   eventtype = UPDATE | CLEAR | RELOAD

   On UPDATE, it adds a new subscription to listen to an
   experiment conveyed by EXPT field of UPDATE message.

   On CLEAR, it removes the subscription for a experiment
   experiment conveyed by EXPT field of CLEAR message.

   On RELOAD, it removes all active experiment subscriptions.

   Also, while it comes up, it sends a RELAOD event to emulab's
   event-server to reload it with all existing experiments:

   host = node's ip addr
   objtype = EVPROXY
   objname = __$nodeid_proxy
   eventtype = RELOAD