Jitter and intrapacket delay

[Carter Bullard]

Hey Urtho,
Remember, the values are in uSec, and jitter is really std deviation,  
rather
than variance, so the number could have been much larger.
With all the loss, yes you can get stdev's that are astronomical.   
Did you
calculate this from a packet file, like I suggested?   With a packet  
file, you
can actually go back and verify the numbers.

Most if not all of them look good, surprisingly good.  When the stdev is
2 orders of magnitude less than the mean (intpkt is mean, jitter is  
stdev)
like it is in your list #4, #5, #7, #8, and #9.  These seem to be the  
same
machines, so maybe they are having the best QoS? If you notice when
there is loss, the jitter really jumps up, which is what its suppose  
to do ;o)

not sure that I understand the NaN in the last one though.  If you
aggregate the data too much, you do end up with overflow conditions
on the sum of squares that we are carrying along.  The value is a
64-bit float, so it should carry a lot of data, but when dealing with
uSec differences, the sum of squares can get pretty high pretty fast.

Carter


ps when you do this this way, try adding a "-L0" to the output to get  
the column
labels, and you can use filters instead of grep.  I know, ..., i  
know, easiest
approach ...., but the filters do help a lot.

On Jul 17, 2007, at 11:31 AM, Urtho wrote:

>
> On 2007-07-16, at 23:19, Carter Bullard wrote:
>
>> Hey Urtho,
>> Ooops, emails went out in the wrong order!!!!
>> I did find a problem with argus() that could have caused the  
>> behavior you
>> were seeing.  I have now uploaded a new argus-3.0.0.tar.gz for you to
>> try out, and if you are having problems after this, please try the  
>> directions
>> below.
>
> Most of the jitter values are off the charts.
> Is this normal ? Can it be because of dropped packets ?
>
> [root at radius argus]#  racluster -p2 -n -r /argus/argus.out -c'|' -s  
> +sintpkt +dintpkt +sjit +djit +dur +sloss +dloss - rtp |grep 16:41
> 16:41:13.802| eI      |rtp|193.192.170.34|5078|<->|62.29.133.121| 
> 5216|10522|12927|778628|956598|CON|24574.00|20000.00|2050588800.00| 
> 1082.00|258.54|2407|0
>
> 17:11:04.348| e       |rtp|193.192.170.34|5112|<->|62.29.133.121| 
> 5300|3007|3005|222518|222370|CON|46458.00|46480.00|1450602.00| 
> 1451297.00|60.12|0|0
> 17:11:16.138| e       |rtp|193.192.170.34|5114|<->|62.29.133.121| 
> 5304|1502|1500|111148|111000|CON|83169.16|83252.92|2447328.00| 
> 2448941.00|30.02|0|0
> 17:12:02.478| e       |rtp|193.192.170.34|5118|<->|62.29.133.121| 
> 5308|5227|5227|386798|386798|CON|20000.00|20000.00|433.00|1363.00| 
> 60.00|0|0
> 17:12:03.979| e    V  |rtp|193.192.170.34|5116|<->|83.15.159.92| 
> 5198|2525|1460|186850|103904|CON|20000.12|33062.74|162.03|63305.26| 
> 50.49|0|1
> 17:12:58.579| e       |rtp|193.192.170.34|5120|<->|62.29.133.121| 
> 5314|1992|1992|147408|147408|CON|56345.86|56366.08|1621761.75| 
> 1622662.50|39.85|0|0
> 17:13:46.270| e       |rtp|193.192.170.34|5122|<->|62.29.133.121| 
> 5322|599|598|44326|44252|CON|106390.70|106603.40|2112601.50| 
> 2116017.25|11.96|0|0
> 17:15:21.408| e       |rtp|193.192.170.34|5124|<->|62.29.133.121| 
> 5326|2232|2231|165168|165094|CON|20000.06|19999.62|226.37|1772.21| 
> 44.62|0|0
> 17:17:22.597| e       |rtp|193.192.170.34|5126|<->|62.29.133.121| 
> 5330|364|362|26936|26788|CON|276862.28|19998.46|4893878.00|692.79| 
> 7.26|0|0
> 17:18:55.026| e       |rtp|193.192.170.34|5130|<->|62.29.133.121| 
> 5338|3007|3005|222518|222370|CON|20000.00|20000.00|24.00|807.00| 
> 60.12|0|0
> 17:27:44.023| eI      |rtp|193.192.170.34|5132|<->|62.29.133.121| 
> 5348|1487|1487|110038|110038|CON|20000.14|20000.08|nan|590.88|29.75| 
> 0|0
>