know what is actually being reported.

Look for the XXX in get_plabinfo to see where it needs to be beefed
up.

Added support for the new buffer-exhaustion method of bandwidth measurement. It appears to do the right thing, but I still need to test it using tg.

instructed via stdin to change its rate of sending and blocksize... and
can also be prescripted.  Unfortunately, I didn't get time to finish the
server...

Added a capped pending buffer which retains write and timing information. Minor reporting fixes. When I checked out start-experiment, it said that there was an RCS version and merged the changes into my version. Therefore checking in this version should be correct. Mike should doublecheck.

corresponding planet-* nodes.

One catch: I only did the backend, I didn't do the part that actually
looks up the plab nodes in the DB.  Once get_plabinfo() is filled in
with a real body, this will work!

in some cases.

Add list of categorized plab nodes.  Note that this list only contains nodes
that were known to be available as of _today_.  I'll need to go back and
add the missing down/unavilable nodes in order for it to be complete.

Changes to bgmon and friends to allow specification of an alternate experiment
to listen/send with (expt event field).  This defaults to "__none".  I also
put in options for specifying bandwidth and latency periods, and generally
fixed up option parsing.

The runbgmon script has also been changed to autodetect the pid/eid and pass
this to bgmon.  You can override this with the "-e" switch.

* opsrecv sends an ack back to bgmon for each packet it receives.

* Rework caching code a bit, and rearrange slightly to deal with acks.

Note that opsrecv does not deal with duplicates yet ... still thinking
about that, but it actually does not matter too much if the info gets
duplicated in the DP; multiple entries with the same data and the same
timestamp *should* be harmless.

 bgmon-running node to that of another node.

The samples with long RTT values are removed for possible lack of symmetric
one-way delays. The remaining samples with offset outliers are also removed
by the median filter.

 * auto-pelab.ns: add an event-group that runs 'update -t' on the nodes
   to update the tarballs/RPMs.  NOTE that you will need to modify your
   experiment and update your version of the NS file for this to work.

 * install-tarball: simple script to run tevc and trigger the tarball
   updates

 * start-experiment: '-p' option to monitor the real planetlab nodes
   instead of the emulated ones, -M/-S to set the arguments passed to
   the monitors/stubs, -m/-s to append to the current arguments passed
   to the monitors/stubs.

 * monitor/auto-monitor.sh: pass through the command line arguments

things like inter-write timings.

to somewhere other than stdout or netmond.

See:     #MARK_RELIABLE  sections

w/o it.

event a chance to get out.

Update the plab proxy to manage restoring the connection to the remote
elvind on ops itself (instead of leaving it up to elvin).

This whole program needs cleanup/refactoring attention.  Later.

It interacts badly with the JVM.

This means we can (and do) report the connect() BEFORE the handshake
is done, which helps keep the stub from starting far behind.

There is a new LocalPort message returned after the connect()
finishes.

Fixed a bug with accept()ed sockets.

results to stdout instead of to netmond.

the monitor and the stubs.  Otherwise the new ones were immediately
being killed off.  Some sort of event delivery issue?

on initial page load; mysqld has gone catatonic and we find out about
it many other ways, so lets not add more email to the flood.

which the delay node needs, gets fired.

Look at "SCHEDULER" attribute to get most everything.