Wireshark-dev: [Wireshark-dev] Question regarding the DELAY
From: "Andreina Toro" <andreina.toro@xxxxxxxxx>
Date: Thu, 26 Oct 2006 11:49:21 -0400
Hi Mr. Ruoff, thanks for helping me with my thesis. What you explained me really helped, I talked to the people here and I´m going to analyze the "Mean Jitter". But I still have a confusion with the delay, you said: "
You can take the average delay over the whole communication for example.
But you should also monitor the maximum delay." I agree with you, and I have some questions about that.
Is the round-trip propagation delay (A - LSR - DLSR) specified in Wireshark??
When I filter RTCP packets I can get specific information about them, I took some fields to ask you some questions...
- Real-time Transport Control Protocol (Sender Report)
Stream setup by H245 (frame 8)
-Source 1
Identifier: 1632266440
SSRC contents
Extended highest sequence number received: 35241
Interarrival jitter: 560
Last SR timestamp: 14635237 (0x00df50e5)
Delay since last SR timestamp: 251396
Frame matching Last SR timestamp: 447
Roundtrip Delay(ms): 5
Is this Roundtrip Delay(ms) the (A - LSR - DLSR) specified in the RFC-3550??.
And why not all the RTCP packets have this Roundtrip Delay? I read in that RFC that if no SR packet has been received yet from SSRC_n, the DLSR field is set to zero.? But why is this?
Last SR timestamp: 0 (0x00000000)
Delay since last SR timestamp: 0
How is the conversion done between "Delay since last SR timestamp: 251396" and the Roundtrip in ms??
Do I calculate the average delay with this Roundtrip Delay data?,
And another question, is there a way to get the maximum delay information with wireshak?
Thanks sooo much for your time!
Best Regards,
Andreina
Lars Ruoff <Lars.Ruoff@xxxxxxxxxx> escribió:
Hello Andreina,
First, the roundtrip delay we are speaking of is not directly the DLSR
field, but A - LSR - DLSR. (see RFC 3550 - page 40/41)
Again it is difficult to define what should be used for a QoS review.
You can take the average delay over the whole communication for example.
But you should also monitor the maximum delay.
best regards,
Lars
Andreina Toro wrote:
> Hi Mr. Ruoff, I was reading the RFC-3550 regarding "delay since last SR
> (DLSR): 32 bits", I know with Wireshark I can filter RTCP so I would be
> able to read this information, but how is the delay managed in a Quality
> of Service Review?. I mean, what has to be less than 150ms? the delay
> since last SR for every packet for every call? :0S How do you clasify a
> call to be one with an acceptable delay?
>
>
> Best regards!
> Andreina
- Prev by Date: Re: [Wireshark-dev] [Wireshark-users] root dispersion in NTP
- Next by Date: [Wireshark-dev] Unicode strings in packets
- Previous by thread: Re: [Wireshark-dev] [Wireshark-users] root dispersion in NTP
- Next by thread: [Wireshark-dev] Unicode strings in packets
- Index(es):