I'm sure almost every VOIP implementation that is using RTP over
UDP for carrying voice traffic does care about their packets arriving
on time ;-) The differences is that timeliness of arrival of those RTP
frames is done through provision of QoS configuration at network
components along the path, usually through attention to queueing and
priority levels, and signalling though protocols like 802.1p and
DiffServ.
Because RTP provides timestamping it makes it a little easier to
quantitatively measure the time-oriented charactistics of that traffic
flow.
Regards, Martin
Martin,
Understood. My point had to do with RTP not relying "exclusively" on
udp-specific features to manage & control RTP traffic, as the
original poster mentioned during subsequent posts.
On Thu, Apr 8, 2010 at 7:17 AM, Kevin
Cullimore <kcullimo@xxxxxxxxxx>
wrote:
On 4/7/2010 3:07 PM, ankur aggarwal wrote:
Hi Anders,
Thanxs for the reply. My problem is bit different. I want to use UDP
only. As http://www.mail-archive.com/wireshark-users@xxxxxxxxxxxxx/msg03818.html
link suggest we cannot use UDP for these calculation, instead use RTP
directly. In my project i need to know the difference in RTP and UDP.
UDP is deployed as a transport protocol precisely when people DON'T
care about measuring traffic that accurately.
Can any body tell me how to convert .pcap (wireshark file) to excel
file. I found this link http://packetbone.com/
. But
unfortunately i donot PC with 32 bit + XP in my lab..
Ankur
On Wed, Apr 7, 2010 at 2:57 PM, Anders
Broman <a.broman@xxxxxxxxx>
wrote:
Hi,
Goggling for VLC player found http://www.videolan.org/vlc/
very breifly looking at the info it looks like there are several options
for how the video stream is sent if it really is RTP in your case you
could use decode as and select RTP and then use the RTP analysis
Note that it designed to analyse Audio and some values may not be
entirely correct/applicable for Video.
You can also use the IO-graph to calculate the bandwidth. If RTP isn't
used you may have to digg trough the documentation/sources
of the VLC player to find out if you can do the calculations you need.
You may have to export the data from the RTP analysis and perform your
own calculation from the raw data or update Wiresharks sources to do it
for you. You may also have to know what codec(s) are used to do correct
calculation. You must also consider where your traces are made and
perhaps compare results
from the sender side with results on the receiver side. No easy answers
here :-)
Regards
Anders
ankur aggarwal skrev 2010-04-07 20:10:
Hi Antony ,
Thanxs for your response. What
does this mean "RTP is commonly carried within UDP"
?? I found that wireshark provide support for RTPas http://wiki.wireshark.org/RTP_statistics
. If i need to calculate jitter,bandwith etc. What should i use. How
can i use UDP
Ankur
On Wed, Apr 7, 2010 at 1:57 PM,
Anthony
Murabito <anthony@xxxxxxxxxxx>
wrote:
Hi Ankur,
RTP is commonly carried within UDP. Although this is not mandatory,
there are other possible underlying protocols such as AAL5.
Regards,
Anthony
On 04/07/2010 01:49 PM, ankur aggarwal wrote:
Hello All,
I am a new user of wireshark. I am working on stream video packets
and receving at other end through wireless hops (Access point). I used
to do this using VLC player and thru UDP protocol. I started using
Wireshark for analysis of my packets. In between i found we cannot
calculate jitter , delay etc.. through UDP. Later i found RTP protocol.
Now the question is what exactly is the difference between UDP and RTP
in term of networking and in term of Wireshark s/w working.
I need some proof as this information will be needed for my
publication.
Help me plz
--
With Regards
Ankur Aggarwal
Research intern
Optical Zeitgeist Laboratory
Institut National de la Recherche Scientifique (INRS) - ÉMT
800, de la Gauchetière Ouest, bureau 6900
Montréal, QC, H5A 1K6
CANADA
Ankur Aggarwal
Research intern
Optical Zeitgeist Laboratory
Institut National de la Recherche Scientifique (INRS) - ÉMT
800, de la Gauchetière Ouest, bureau 6900
Montréal, QC, H5A 1K6
CANADA
Ankur Aggarwal
Research intern
Optical Zeitgeist Laboratory
Institut National de la Recherche Scientifique (INRS) - ÉMT
800, de la Gauchetière Ouest, bureau 6900
Montréal, QC, H5A 1K6
CANADA