Reduce CPU so that iperf client is CPU bound

If we want to observe a monotonic affect, we should
make sure that we are in fact CPU limited where it
matters. In this case, we are CPU limiting the hosts,
and the iperf client uses a lot of CPU. We need to
reduce the CPU allocation so that iperf is in fact
CPU bound.

We also correct the CPU allocation so that the client
and server each receive 50% of the total. Previously
we were specifying the per-host CPU allocation, so
45% meant we were allocating 90% of the overall CPU,
which seems a bit confusing. On the other hand, now
at 40% each host gets 20% of the CPU, which could also
be considered slightly confusing!

Although the client transmit rate is going to be the
limiting factor, we still measure the received data
rate at the server, because that is more interesting
than the initial burst of buffering at the client.
Measuring at the server becomes more important as
we reduce the iperf time.

The output is also changed slightly, and the test has
been updated appropriately.

examples/test/test_cpu.py

@@ -5,13 +5,12 @@ Test for cpu.py
 
 results format:
 
-    sched   cpu client MB/s
-
-    cfs 45.00%  13254.669841
-    cfs 40.00%  11822.441399
-    cfs 30.00%  5112.963009
-    cfs 20.00%  3449.090009
-    cfs 10.00%  2271.741564
+sched	cpu	received bits/sec
+cfs	50%	8.14e+09
+cfs	40%	6.48e+09
+cfs	30%	4.56e+09
+cfs	20%	2.84e+09
+cfs	10%	1.29e+09
 
 """
 
@@ -28,7 +27,7 @@ class testCPU( unittest.TestCase ):
         "Verify that CPU utilization is monotonically decreasing for each scheduler"
         p = pexpect.spawn( 'python -m mininet.examples.cpu' )
         # matches each line from results( shown above )
-        opts = [ '([a-z]+)\t([\d\.]+)%\t([\d\.]+)',
+        opts = [ '([a-z]+)\t([\d\.]+)%\t([\d\.e\+]+)',
                  pexpect.EOF ]
         scheds = []
         while True: