Wireshark mailing list archives

Re: Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏


From: "RUOFF, LARS (LARS)** CTR **" <lars.ruoff () alcatel-lucent com>
Date: Mon, 26 Sep 2011 14:49:24 +0200


Let's say that RFC3550 (RTP) explicitly allows for negative counts:

   cumulative number of packets lost: 24 bits
      The total number of RTP data packets from source SSRC_n that have
      been lost since the beginning of reception.  This number is
      defined to be the number of packets expected less the number of
      packets actually received, where the number of packets received
      includes any which are late or duplicates.  Thus, packets that
      arrive late are not counted as lost, and the loss may be negative
      if there are duplicates.  The number of packets expected is
      defined to be the extended last sequence number received, as
      defined next, less the initial sequence number received.  This may
      be calculated as shown in Appendix A.3.

...and Wireshark applies this by the letter in RTP analysis.

Regards,
Lars
 

________________________________

From: Farooq Razzaque [mailto:farooq_mcp () hotmail com] 
Sent: lundi 26 septembre 2011 14:36
To: wireshark-users () wireshark org; RUOFF, LARS (LARS)** CTR **
Subject: RE: [Wireshark-users] Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏


Dear Lars
 
Really thanks for your support.
 
Can u please let me know that is this the wireshark or RTP behaviour of showing the duplicate packets in negative value 



<http://www.flamingtext.com/hmail.html> 


 

  

From: lars.ruoff () alcatel-lucent com
To: wireshark-users () wireshark org
Date: Mon, 26 Sep 2011 13:56:49 +0200
Subject: Re: [Wireshark-users] Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏

Farooq,
(I put this back on the list if you don't mind, so others can comment and it gets archived.)

In reality, there were only 1744/4 = 436 unique RTP packets sent between the endpoints.
But Wireshark captured each packet 4 times.
(Note that your packet counts are always multiples of 4)
Each duplicate packet (packet with same RTP sequence number) gives rise to a lost count of -1.
Thus from the 1744, 436 were unique, the remaing 3/4 i.e. 1308 are duplicate.
Thus a lost count of -1308, corresponding to -300% of the 100% unique packets.
Hope this is clear.


regards,
Lars


________________________________> 
From: Farooq Razzaque [mailto:farooq_mcp () hotmail com] 
Sent: lundi 26 septembre 2011 12:21
To: jaap.keuter () xs4all nl; RUOFF, LARS (LARS)** CTR **
Subject: RE: [Wireshark-users] Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏


Dear Lars/Jaap

Thanks for your support.

we are SPANing the data on cisco switch and forwarding to Alcatel and Cisco recording machine.

monitor session 2 source interface x
monitor session2 destination interface x.

Can u please let me know how the following can be seen by analysis engine by 4 times. how it is calculate

Number of packet = 1744
Lost -1308 (-300)



<http://www.flamingtext.com/hmail.html> 






From: lars.ruoff () alcatel-lucent com
To: wireshark-users@w! ireshark.org
Date: Mon, 26 Sep 2011 09:44:30 +0200
; > Subject: Re: [Wireshark-users] Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏

Hi,

No, since you (almost) consistently have -300% all the time, it is most likely that every packet has been seen 
exactly 4 times by the analysis engine, but no packets have been lost.
(It is an artefact of the RFC3550 lost packets algorithm that duplicate packets are counted as negative losses)
However, as Jaap noted, in order to get more readable data, you should fix your capture setup issue which makes you 
see every packet multiple times.

Regards,
Lars



________________________________

From: wireshark-users-bounces () wireshark org [mailto:wireshark-users-bounces () wireshark org] On Behalf Of 
Farooq Razzaque
Sent: same! di 24 septembre 2011 19:04
To: wire! shark-users () wireshark org
Subject: [Wireshark-users] Wireshark RTP Stream - Packet Lost in Neg value over the WAN‏


Dear All




Can u have a look at the attached screen shot of wireshark. In LOST COLUMN it is showing 300% , -299.7% pack lost. 



Do u have any idea that are these packet loss is normal/abnormal.



IP phones ( 172.20.24.x) are located in one branch and Recording machine (172.20.19.17) is located in other branch.



SPANing is happing over the WAN via L2TPV3.



IP Phones : 172.20.24.X (IP Phone)



172.20.19.17 (Recording machine) 

____________________________! _______________________________________________
Sent via: Wireshark-users mailing list <wireshark-use! rs () wireshark org>
Archives: http://www.wireshark.org/lists/ wireshark-users
Unsubscribe: https://wireshark.org/mailman/options/wireshark-users
mailto:wireshark-users-request () wireshark org?subject=unsubscribe

___________________________________________________________________________
Sent via: Wireshark-users mailing list <wireshark-users () wireshark org>
Archives: http://www.wireshark.org/lists/wireshark-users
Unsubscribe: https://wireshark.org/mailman/options/wireshark-users
mailto:wireshark-users-request () wireshark org?subject=unsubscribe

___________________________________________________________________________
Sent via:    Wireshark-users mailing list <wireshark-users () wireshark org>
Archives:    http://www.wireshark.org/lists/wireshark-users
Unsubscribe: https://wireshark.org/mailman/options/wireshark-users
             mailto:wireshark-users-request () wireshark org?subject=unsubscribe

Current thread: