Understanding of Cisco SAA (Service Assurance Agent)/RTR for VOIP

[josephn]

Hi,

I need some detailed information for the following variables captured as reflected as statistics on the voip gateway (cisco AS5300):-

1.MinPositiveSD

2. MaxPositiveSD

3. NumofPositiveSD

4. SumofPositiveSD

We had configured the probe and responder in which the probe will send out about 600 ICMP packets in 1 hour (10 packets/minute)to the responder with the delay of 10 ms between two packets.

My question here is how will one derive or calculate the positive jitter effectively by using the above variables. Please refer to the following example :-

RTT Values:

NumOfRTT: 600 RTTAvg: 234 RTTMin: 231 RTTMax: 251

RTTSum: 140487 RTTSum2: 32903363

Packet Loss Values:

PacketLossSD: 0 PacketLossDS: 0

PacketOutOfSequence: 0 PacketMIA: 0 PacketLateArrival: 0

InternalError: 0 Busies: 0

Jitter Values:

NumOfJitterSamples: 540

MinOfPositivesSD: 1 MaxOfPositivesSD: 8

NumOfPositivesSD: 75 SumOfPositivesSD: 108 Sum2PositivesSD: 280

MinOfNegativesSD: 1 MaxOfNegativesSD: 10

NumOfNegativesSD: 80 SumOfNegativesSD: 114 Sum2NegativesSD: 310

Positive Jitter = 108/75 = 1.44 ms

Negative Jitter = 114/80 = 1.425 ms

Is the above calculation correct ? if yes, what does MaxofPositiveSD = 8ms tells us. Which parameter should be used as threshold value for voip monitoring 1.44ms or 8 ms ?

Thanks.

regards,

Joseph

[grayd]

Your 1.44 ms calculation would be the average jitter. The 8 ms would be the maximum jitter. Which one is important depends on what you want to see. Jitter is a packet by packet measure. Unfortunately, this means that one delayed packet will throw off both your max pos and max neg (delayed packet takes a lot longer to get there, the next normal packet takes a lot less time to get there than the delayed packet). However, I think that with VoIP, the worst case scenario is the one to measure so I normally look at that value more than the average, but I track both.