All things related to multicast

[Philip Pishioneri <>]

Alan Crosswell wrote:

> So I am trying to view some sources and have been seeing the following
> very consistently:
> ...
> Could others possibly view these same sessions and compare?
> Is there a better way than these subjective measurements to compare these?


I found an old-ish ('93) program called "rtpqual" (from Matt Mathis) 
which gives some objective measurements.  A couple of places which had it:

  ftp.ee.lbl.gov -- (don't have exact location)
  http://www.mit.edu/afs/sipb/project/sipbnet/src/rtpqual/rtpqual.c

(I'm using the "sipb" version.  Had tried to find it once at psc without 
success.)

It takes the multicast group and port and gives output like (just did 
this for "U0 Broadcasts NASA Videos (01)" video stream):

Defaulting to: ./rtpqual 224.2.196.83 62002 rtp
T Pkts Loss %  Late Bytes |  Pkts Loss   %    Late     kB Sender
20  99  3  2.00  9 117123 |    99    3  2.00    9     114 128.223.230.9
21  77  0  0.00  9  89843 |   176    3  1.00   18     202 128.223.230.9
22  98  1  1.00  9 112451 |   274    4  1.00   27     311 128.223.230.9
23 101  0  0.00  9 116248 |   375    4  1.00   36     425 128.223.230.9
24  75  2  2.00  9  85543 |   450    6  1.00   45     508 128.223.230.9
25 103  0  0.00  9 117456 |   553    6  1.00   54     623 128.223.230.9
26  76  1  1.00  9  84927 |   629    7  1.00   63     706 128.223.230.9
27 103  0  0.00  9 116856 |   732    7  0.00   72     820 128.223.230.9
28 101  0  0.00  9 114968 |   833    7  0.00   81     933 128.223.230.9
29 101  0  0.00  9 112408 |   934    7  0.00   90    1042 128.223.230.9
30  81  0  0.00  9  91256 |  1015    7  0.00   99    1131 128.223.230.9
31  97  3  3.00  9 105836 |  1112   10  0.00  108    1235 128.223.230.9
32  93  0  0.00  9  96861 |  1205   10  0.00  117    1329 128.223.230.9
33 111  0  0.00  9 116957 |  1316   10  0.00  126    1444 128.223.230.9
34 107  1  0.00  9 115111 |  1423   11  0.00  135    1556 128.223.230.9
35  79  0  0.00  9  88008 |  1502   11  0.00  144    1642 128.223.230.9
36 105  0  0.00  9 116744 |  1607   11  0.00  153    1756 128.223.230.9
37  79  0  0.00  9  87994 |  1686   11  0.00  162    1842 128.223.230.9
38 106  0  0.00  9 115694 |  1792   11  0.00  171    1955 128.223.230.9
39  76  0  0.00  9  87708 |  1868   11  0.00  180    2041 128.223.230.9
40 103  1  0.00  9 115687 |  1971   12  0.00  189    2153 128.223.230.9
41  77  0  0.00  9  86815 |  2048   12  0.00  198    2238 128.223.230.9
42  79  0  0.00  9  87603 |  2127   12  0.00  207    2324 128.223.230.9
43 109  0  0.00  9 116752 |  2236   12  0.00  216    2438 128.223.230.9

Comments from the program (one source file):

> Each output row presents the statistics from one transmitter.
>
> Column one is the seconds part of the system time
>         such that loss events can be correlated with other events.
> Statistics from the current second are displayed in columns 2-6:
>         data packets received,
>         (presumed) lost data packets,
>         percentage loss,
>         late data (and any non-sequenced control) packets,
>           (Currently only checks for out of sequence, but not excessive delay)
>         bytes received (data, late data and control).
> The next 5 columns are cumulative totals for this transmitter
>         (except k Bytes instead of Bytes).
>
> If there is more than one transmitter, cumulative total statistics are
> displayed for each every minute.
>
> BUGS: Delay variance is not instrumented.
>
> Each line appears only when there is input in a new second, therefor it is not
> possible to see the last second of a transmission.


-Phil