[Home]

Summary:ASTERISK-24584: RTP QOS Channel variables not being set correctly
Reporter:Taylor Hawkes (antiochIst)Labels:
Date Opened:2014-12-02 15:47:41.000-0600Date Closed:2021-09-17 16:38:12
Priority:MajorRegression?
Status:Closed/CompleteComponents:Core/RTP
Versions:12.7.0 13.0.0 Frequency of
Occurrence		Constant
Related
Issues:
is duplicated byASTERISK-26710 [patch] res_rtp_asterisk: CHANNEL arguments, (rtcp,all_rtt),(rtcp,all_loss),(rtcp,all_jitter) always return 0
is related toASTERISK-14745 rtt should be stored as double in struct ast_rtp_instance_stats
Environment:Attachments:( 0) ASTERISK-24584.diff
Description:version 11+ QOS channel vars are all 0.0000 except for (RTPAUDIOQOS).

klaus3000 noticed this a while ago and pointed out bug & fix in this issue: ASTERISK-14745

After applying the suggested fix, the values are at least being set as show here:
{noformat}
>RTPAUDIOQOSRTTBRIDGED=minrtt=0.000000;maxrtt=0.000000;avgrtt=0.000000;stdevrtt=0.000000;
   >RTPAUDIOQOSLOSSBRIDGED=minrxlost=65351.000000;maxrxlost=65351.000000;avgrxlost=65351.000000;stdevrxlost=65351.000000;reported_minlost=0.000000;reported_maxlost=0.000000;reported_avglost=0.000000;reported_stdevlost=0.000000;
     >RTPAUDIOQOSJITTERBRIDGED=minrxjitter=0.000002;maxrxjitter=0.000054;avgrxjitter=0.000036;stdevrxjitter=inf;reported_minjitter=0.000000;reported_maxjitter=0.000000;reported_avgjitter=0.000000;reported_stdevjitter=0.000000;
      >RTPAUDIOQOSBRIDGED=ssrc=929865227;themssrc=4181318442;lp=65351;rxjitter=0.000000;rxcount=370;txjitter=0.000035;txcount=373;rlp=0;rtt=0.000000
      >RTPAUDIOQOSRTT=minrtt=4293.621143;maxrtt=4294.705748;avgrtt=4294.163446;stdevrtt=18444498.197134;
      >RTPAUDIOQOSLOSS=minrxlost=1.000000;maxrxlost=1.000000;avgrxlost=1.000000;stdevrxlost=1.000000;reported_minlost=0.000000;reported_maxlost=0.000000;reported_avglost=0.000000;reported_stdevlost=0.000000;
      >RTPAUDIOQOSJITTER=minrxjitter=0.000589;maxrxjitter=0.059641;avgrxjitter=0.008099;stdevrxjitter=inf;reported_minjitter=0.000000;reported_maxjitter=10.000000;reported_avgjitter=5.000000;reported_stdevjitter=11.180340;
      >RTPAUDIOQOS=ssrc=176689378;themssrc=4025103005;lp=2;rxjitter=0.000000;rxcount=502;txjitter=0.003975;txcount=347;rlp=0;rtt=4293.621143
{noformat}

Before the fix they look like this:
{noformat}
>RTPAUDIOQOSRTTBRIDGED=minrtt=0.000000;maxrtt=0.000000;avgrtt=0.000000;stdevrtt=0.000000;
      >RTPAUDIOQOSLOSSBRIDGED=minrxlost=0.000000;maxrxlost=0.000000;avgrxlost=0.000000;stdevrxlost=0.000000;reported_minlost=0.000000;reported_maxlost=0.000000;reported_avglost=0.000000;reported_stdevlost=0.000000;
      >RTPAUDIOQOSJITTERBRIDGED=minrxjitter=0.000000;maxrxjitter=0.000000;avgrxjitter=0.000000;stdevrxjitter=0.000000;reported_minjitter=0.000000;reported_maxjitter=0.000000;reported_avgjitter=0.000000;reported_stdevjitter=0.000000;
      >RTPAUDIOQOSBRIDGED=ssrc=2145217062;themssrc=72888040;lp=65325;rxjitter=0.000000;rxcount=1245;txjitter=0.000044;txcount=1243;rlp=0;rtt=0.000000
      >RTPAUDIOQOSRTT=minrtt=0.000000;maxrtt=0.000000;avgrtt=0.000000;stdevrtt=0.000000;
      >RTPAUDIOQOSLOSS=minrxlost=0.000000;maxrxlost=0.000000;avgrxlost=0.000000;stdevrxlost=0.000000;reported_minlost=0.000000;reported_maxlost=0.000000;reported_avglost=0.000000;reported_stdevlost=0.000000;
      >RTPAUDIOQOSJITTER=minrxjitter=0.000000;maxrxjitter=0.000000;avgrxjitter=0.000000;stdevrxjitter=0.000000;reported_minjitter=0.000000;reported_maxjitter=0.000000;reported_avgjitter=0.000000;reported_stdevjitter=0.000000;
      >RTPAUDIOQOS=ssrc=1525160020;themssrc=2865122643;lp=11;rxjitter=0.000000;rxcount=1346;txjitter=0.003079;txcount=1233;rlp=0;rtt=4294.602330
{noformat}
Comments:By: Matt Jordan (mjordan) 2014-12-03 09:53:05.470-0600

Are these numbers actually correct however? As an example, this looks a bit odd:

{noformat}
   >RTPAUDIOQOSLOSSBRIDGED=minrxlost=65351.000000;maxrxlost=65351.000000;avgrxlost=65351.000000;stdevrxlost=65351.000000;reported_minlost=0.000000;reported_maxlost=0.000000;reported_avglost=0.000000;reported_stdevlost=0.000000;
{noformat}

# 65361 seems a bit high
# The average/stddev seem a bit odd (they equal the min/max)


By: Taylor Hawkes (antiochIst) 2014-12-03 10:02:27.376-0600

No, those numbers are still not correct.  The above patch only fixes the setting of the channel vars - the tricky part is still in need of fixing.


By: Rusty Newton (rnewton) 2014-12-04 16:02:59.211-0600

Taylor, do you want to submit a patch yourself so that we will have something on this issue? Or do you want to contact Klaus and see if he wants to submit a patch here?


By: Jason Marble (marble) 2016-02-16 12:43:55.031-0600

Taylor, have you or do you know of anyone that's made progress on this?
By: newborn (newborn838) 2016-07-08 08:18:43.047-0500

Bump
Same problem here
Anyway, does anybody know how to catch RTP SSRC value (0x....) ?
This can be useful to dump only current call via tshark.
By: Richard Mudgett (rmudgett) 2017-03-28 13:42:19.241-0500

Issue was actually fixed by ASTERISK-26710
By: Richard Mudgett (rmudgett) 2017-03-28 13:49:29.377-0500

A version of the initial patch was committed by ASTERISK-26710 but apparently the values are still wrong.
By: Taylor Hawkes (antiochIst) 2017-03-28 19:01:38.530-0500

Rusty,  Jason - Fixing this is out of my level of expertise, will pry need someone who knows what there doing to take a look.  
By: Sean Bright (seanbright) 2021-09-17 16:38:12.585-0500

Fixed by https://gerrit.asterisk.org/c/asterisk/+/4707