Zach Gibbs, Content Developer, Education Services, Juniper Networks 

Paragon Active Assurance — VoIP UDP Test

Learning Bytes
Zach Gibbs Headshot
Slide with a headline that says, “Paragon Active Assurance — VoIP UDP Test.” Below that, on the left, is a network diagram with a cloud filled with the words, “Service Provider.” Below that there are two icons of suns with the words “eth1,” “eth0,” and “TA-1” around them. Into between is a smaller cloud with the word, “Mgmt.” Below that is an icon with the words “Control Center.” On the right is a bullet list that says:   What are we doing to do?   Configure VoIP UDP task   Duration: 30 seconds   Setup Type: Full-Mesh   Codec: GSM EFR   Port: 1001   ES threshold   MOS: 5   Expected DSCP: 10   Run the test and examine the results

Juniper Learning Byte: Run a VOIP UDP Test the easy way.

Tune into this Learning Byte from Juniper’s Zach Gibbs to learn step by step how to configure the VoIP UDP task using the Control Center Web Interface. This video is most appropriate for users who are new to working with Paragon Active Assurance. 

Show more

You’ll learn

  • The devices involved in the topology for configuring VoIP UDP talk 

  • How the control center communicates with the devices 

  • How to run the test and examine the results 

Who is this for?

Network Professionals Business Leaders


Zach Gibbs Headshot
Zach Gibbs
Content Developer, Education Services, Juniper Networks 


00:00 [Music]

00:13 hello my name is zach gibbs and i'm a

00:16 content developer within education

00:18 services inside juniper networks and

00:21 today we will be going through the

00:22 paragon active assurance voip udp test

00:26 learning bike

00:28 alright so here is our topology we have

00:30 three devices that we're concerned about

00:32 in the topology that is ta1 ta2 and the

00:35 control center ta1 and ta2 use eth0 to

00:39 connect into the management network and

00:41 then the control center can communicate

00:43 with those devices through the

00:45 management network and then ta1 and ta2

00:48 use the eth1 interfaces to connect into

00:50 the service provider production network

00:52 all right so with that what are we going

00:54 to do well we're going to configure the

00:57 voip udp task we're going to set the

00:59 duration to 30 seconds the setup type is

01:01 going to be full mesh the codec is going

01:03 to be gsm efr the port is going to be

01:06 1001 and the es thresholds are going to

01:08 be a mos score of 5 and an expected dscp

01:12 value of 10. so one thing i do want to

01:14 talk about is what this task is all

01:16 about and what it does is it simulates

01:18 voip media flows using udp as the

01:21 transport protocol and so these flows uh

01:25 use a specific voice codec it's either

01:27 going to be the

01:28 g.711 the g.723

01:31 the g.729 or the gsm efr codec

01:36 and so with this we can run this task as

01:38 a hub and spoke topology or a full mesh

01:40 topology and in this learning bite we're

01:42 going to be doing the full mesh topology

01:44 and so what this task basically allows

01:46 you to see is the effect your network

01:49 has on voip traffic so basically is your

01:52 network good enough to run voip traffic

01:54 and so to do this there's the mos score

01:58 which stands for mean opinion scale and

02:00 so

02:01 how that works is the mos score is

02:04 calculated by the task

02:06 using the jitter and pack loss of the

02:08 flow

02:09 now the small score can range from five

02:12 being the best and one being the

02:15 absolute worst if you if you get a one

02:17 with the mos score it is definitely your

02:18 network is definitely not ready for voip

02:20 traffic if you get a five it's going to

02:22 be crystal clear uh anything in between

02:25 like a four is great three not so great

02:28 and so forth

02:30 all right so with that we're going to

02:32 set up a mos score five now a mos score

02:34 five means it's it's absolutely perfect

02:36 so we're gonna see what happens with

02:38 that when we set that score uh something

02:40 else to point out is that when the task

02:41 starts the test agents they generate

02:43 traffic at a fixed bit rate and frame

02:45 size

02:46 and that's dependent on the codec that

02:48 you choose for example if we choose

02:51 g.711 for the codec the bitrate is 87.2

02:56 kilobits per second and the frame size

02:58 is 218 bytes

03:01 and so it just depends on whatever codec

03:03 we use okay so with that being said

03:05 let's go ahead and jump to the web

03:06 interface for the control center and get

03:08 this going

03:10 all right so here is the web interface

03:13 for the control center let's go to tests

03:15 and set up a new test sequence we'll

03:18 call this voip

03:20 udp lb that'll be short for learning

03:22 byte and we'll select the voip udp task

03:27 and notice how this is a puzzle piece

03:28 test which means we can add other puzzle

03:30 pieces into the same step and run

03:33 those tasks as a parallel task so keep

03:36 that in mind right so we're going to set

03:38 the duration of 30 seconds

03:40 we're going to set the setup type to

03:42 full mesh

03:43 then we'll select our clients that'll be

03:44 ta1 on eth1 and ta2 on e2 want to make

03:47 sure you select the right interfaces if

03:50 we were to select ta1 eth1 and t82 0

03:53 those are on two separate networks they

03:54 will not be able to communicate with

03:56 each other now we're going to have a bad

03:57 time if we do that so just make sure you

03:59 select the right test agents and the

04:00 right interfaces all right the number of

04:03 flows we didn't specify that in our

04:04 criteria but one works great

04:07 the codec this is where we select our

04:08 codec we want to select the gsm efr

04:11 codec i want to set the port to 1001

04:15 and then we have the threshold for aired

04:17 seconds

04:18 now this is where we set the mos

04:20 value here we're setting to 4 by default

04:23 remember that we want to set that to 5.

04:24 so we're basically saying it's got to be

04:26 near perfect for us to accept this right

04:29 and so set the dscp value to 10 the

04:32 expected dscp value that is and then we

04:35 can go ahead and start the test and

04:37 let's see what happens all right so the

04:38 test is starting up take just a second

04:41 to start getting some results and

04:42 whatnot but we'll see what we get and

04:45 we're already getting some results and

04:46 we're currently getting aired seconds

04:48 this doesn't look good it's uh you know

04:50 we want to see green here we're seeing

04:52 black and so you gotta ask yourself well

04:54 what's the reason why am i seeing aired

04:55 seconds in the air second history bar

04:58 and

04:59 why are things not green and the reason

05:01 behind that is recall that we set the

05:03 mos score to five which is as you know

05:06 perfect or just excellent quality that's

05:09 gonna be really hard to achieve right

05:10 and you can see here the mos score is

05:12 set to 4.3 or the mos score result for

05:14 the streams both streams is 4.3

05:17 and that's not five that's less than

05:19 five that's still really excellent but

05:21 it's not five and let's look at the

05:23 actual stream itself

05:25 to see this more in action get rid of

05:27 the es bar and the rate

05:29 and we see the average delay we select

05:31 the jitter

05:32 get a little rid of delay there as well

05:34 and you can see the mos score which is

05:36 up here at just around just floats

05:39 around 4.3 we actually got up to

05:41 4.51 so it did go up a little bit

05:44 and so you can see that we have some

05:46 jader here that

05:47 the uh

05:48 task isn't liking it's not giving us

05:50 that perfect

05:51 five mos score and we can see that in

05:54 the table output with the results as

05:55 well and we can see that we have the mos

05:58 score 2.9 2.9 out definitely less than

06:01 five or 4.29 excuse me not 2.9 uh 4.29

06:07 and 2.9 would be terrible 4.29 and we

06:10 see that the es dcp

06:13 uh

06:13 percent like we're not having any sort

06:15 of problems there so that looks good

06:16 there so it's definitely the mos score

06:19 is lower than five

06:20 and so let's go ahead and cheat a little

06:22 and just click rerun and change the mos

06:24 score to four right that's what you do

06:26 if

06:27 if you're not passing your test you just

06:28 lower the pass score right no i'm just

06:30 kidding of course

06:32 but i just want to show you what it

06:33 looks like when we set it to four and

06:36 everything uh looks good because we'll

06:37 get a round of 4.3 which is above four

06:40 now remember four is actually really

06:42 really good so it's nothing to be

06:43 embarrassed about if you're getting four

06:44 in your network your voip traffic will

06:46 sound excellent with a 4.

06:48 and so

06:50 let's go ahead and start this

06:53 and this test of course will start in

06:55 just a second and you can see here that

06:57 we're not getting errored seconds in the

06:58 aired history bar or aired seconds

07:00 history bar and we have the mos score

07:02 4.3

07:04 and so things are looking good this is

07:06 what we want to see

07:07 we're passing

07:09 with this test with the last test we

07:10 didn't because we had the mos score set

07:12 at five instead of four here with the

07:14 mos score set of four the test is

07:17 definitely uh resulting in a pass

07:21 so this does bring us to the end of this

07:23 learning byte and in this learning byte

07:25 we demonstrated how to configure and run

07:26 a voip udp test so as always thanks for

07:29 watching

07:32 visit the juniper education services

07:35 website to learn more about courses

07:38 view our full range of classroom online

07:41 and e-learning courses

07:43 learning paths

07:44 industry segment and technology specific

07:47 training paths

07:49 juniper networks certification program

07:51 the ultimate demonstration of your

07:53 competence and the training community

07:56 from forums to social media join the

07:59 discussion

Show more