cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
383
Views
4
Helpful
2
Replies

help with AS5300 Packet stats

brendan.oneill
Level 1
Level 1

I am trying to understand some QoV stats reported in the 200 OK response to a MGCP DLCX message.

Referring to the example below can somebody please answer the following questions.

- is packet loss(P:PL) the sum of lost packets in BOTH directions?

- How is jitter(P:JI) and latency(P:LA) calculated? are they the average for the call duration?

- Why is P:PS different to DSP:TX?

-Is there any good documentation that explains these values?

Thanks in advance.

MGCP...... -> 250 19144821 OK

P: PS=2725, OS=218000, PR=2662, OR=212960, PL=0, JI=24, LA=4

DSP/TX: PK=2701, SG=0, NS=0, DU=54030, VO=54030

DSP/RX: PK=2638, SG=0, CF=0, RX=52730, VO=52680, BS=0, BP=1, LP=0

DSP/PD: CU=24, MI=24, MA=65, CO=1355428424, IJ=1

DSP/PE: PC=0, IC=0, SC=0, RM=60, BO=0, EE=0

DSP/LE: TP=3, TX=0, RP=-66, RM=-664, BN=0, ER=20, AC=20, TA=3,RA=-66

DSP/ER: RD=0, TD=0, RC=0, TC=0

DSP/IC: IC=0

2 Replies 2

Chester Rieman
Cisco Employee
Cisco Employee

These should be stats for the RTP that is terminating on the 5300. Packet loss is total packet loss for one direction. Jitter is a dynamic calculation, this is average jitter for the entire call. Latency is calculated using the RTP timestamps in the packets.

The DSP/PD line displays more jitter info.

MI=minimum de-jitter buffer depth, MA=max.

The thing to watch here for jitter is the BO(buffer overflow)..these are packets that get dropped due to too much jitter. Also watch PC, IC and SC(predictive conceal, interpolate conceal and silence conceal) these would indicate early/late/missing packets and/or jitter.

The reason for the difference between P:PS and DSP/TX is that it take a few milliseconds for the DSP and the IOS voice routine to synce up...a few packets may be lost here.

http://cisco.com/en/US/products/sw/iosswrel/ps5187/products_feature_guide09186a00801bb4d2.html

Hope this helps....

Thanks.

This was very helpful.