B Appendix: Monitoring Practices
This Appendix summarizes a few useful monitoring practices.
B.1 RTP Monitoring
When running video inside an RTP wrapper it is possible to exactly deduce the number of dropped IP
frames due to network issues. This is possible as a result of the 16-bit sequence counter inside the RTP
header. When the protocol mapping is nTS/RTP the RTP parameters
RTPdrop
,
RTPdup
,
RTPooo
and
RTPlag
will be updated and the corresponding alarms
Packet drops:N
,
Duplicate packets:N
and
Out
of order packets(lag:N) are fired (if not switched off).
Note that the probe will perform out-of-order corrections before RTP packet loss analysis is performed.
Example of RTP sequences and their effects on monitoring:
Sequence
Effect
. . . , 10, 11, 12, 13, 14, 17, 18, 19, . . .
2 dropped packets (15-16)
Monitoring page: RTPdrop:+2
Alarms & events: RTP Packet drop: 2
. . . , 10, 12, 13, 16, 17, 18, 19, . . .
1 and 2 dropped packets (11, 14-15)
Monitoring page: RTPdrop:+3
Alarms & events: RTP Packet drop: 3
. . . , 10, 11, 15, 12, 14, 16, 18, 19, . . .
2 dropped packets (13, 17)
1 out of order packets of order 3 (15− >12)
Monitoring page: RTPdrop:+2
Monitoring page: RTPooo:+1
Monitoring page: RTPlag: 3 (at least)
Alarms & events: RTP Packet drops: 2
Alarms & events: RTP out of order packets (lag:3)
B.2 Default Multicast Monitoring
When the protocol mapping is nTS/UDP, meaning there is no RTP information in the multicast stream,
there is no easy way to isolate and register network-induced errors. Assumptions can be done by
performing continuity counter analysis for the content of each received UDP-frame on the fly. The probe
will note CC-errors (CCerr) and generate corresponding alarms (CC skips:N).
VB2xx GigE User’s Manual version 5.4 253
To Next Page
Loading...
Need help?
General
