Paragon Active Assurance — VoIP UDP Monitor

00:00 [Music]

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

00:16 content developer within education

00:18 services inside juniper networks

00:21 and today we will be going through the

00:23 paragon active assurance voip udp

00:26 monitor learning byte

00:29 all right so here is our topology in

00:31 this topology we have three main devices

00:33 that we are concerned about we have ta1

00:36 and ta2 which are test agents and then

00:38 we have the control center and the

00:40 control center communicates with both

00:43 test agents over the management network

00:45 and both those test agents use eth0 to

00:47 connect to the management network and

00:49 then ta1 and ta2 use eth1 to connect

00:53 into the service provider production

00:55 network and so with that being said what

00:57 are we trying to do with this learning

00:58 braid well we want to configure a voip

01:01 udp task since this is a monitor the

01:04 duration is going to be indefinite

01:05 that's just how monitors run the setup

01:08 type is going to be full mesh the codec

01:10 we are going to be using is the gsm efr

01:12 codec we're going to be using port 1001

01:15 and for error seconds threshold we want

01:18 to set the mos value to 5 and the

01:20 expected dcp value to 10. now with this

01:23 task what it does is it simulates

01:26 voip media flows using you guessed it

01:29 udp hence the name voip udp for the task

01:32 and so these flows use a specific voice

01:35 codec that's either g

01:37 711 g.723 g.729 or gsm efr and in this

01:43 learning byte we will be using the gsm

01:45 efr coding and so we can run this task

01:48 as a hub and spoke topology or as a full

01:50 mesh topology we're going to be using

01:52 full mesh for this learning bite and so

01:53 this task basically allows you to see

01:55 the effect your network has on voip

01:57 traffic is my network going to work with

02:00 voip traffic is what you have to ask

02:01 yourself here and this is what this

02:02 helps you do and so to do this we have a

02:05 mos scale and what that stands for is

02:07 mean opinion scale and this mos score is

02:10 calculated by the task by using the

02:12 jitter and packet loss of the flow and

02:14 so if you have high jitter high packet

02:15 loss you're going to get a lower mos

02:17 score now with that

02:19 the mos score can range from five to one

02:22 one being the worst five being the best

02:25 and notice how for this learning byte

02:26 we're going to set a mos score of five

02:29 and see how that is now one thing i want

02:31 to point out it's actually really kind

02:32 of hard to achieve a mos score of five

02:34 and so if we don't get a mos score five

02:36 we'll drop it down just so we can see

02:38 what it looks like uh when it fails and

02:40 also when it passes okay so when this

02:43 test starts what happens is the test

02:45 agents will generate traffic

02:47 at a fixed rate and also frame size the

02:49 rate and the frame size is based on what

02:52 codec we're going to use for example if

02:53 we chose the g.711 uh for the codec the

02:57 bit rate is going to be 87.2 kilobits

03:00 per second and the frame size will be

03:02 218. so 218 bytes

03:05 so keep that in mind how this works it's

03:06 based on the codec we used and then with

03:08 task we're going to set the mos score we

03:10 want

03:11 and see if it actually passes

03:13 and so with that let's go ahead and jump

03:16 to the web interface for control center

03:18 and get this going

03:20 all right so here is the web interface

03:22 for the control center let's go to

03:24 monitoring new monitor

03:27 and we'll create this monitor we'll call

03:28 this

03:29 voip udp lb

03:32 for learning byte and then we'll select

03:34 the voip udp task

03:37 and notice how this is a puzzle piece

03:39 task which means we can add additional

03:41 tasks for this monitor but for this

03:43 learning byte we're just going to focus

03:45 on the voip udp task

03:47 okay so with that setup type full mesh

03:50 and the clients we're going to say ta1

03:52 eth1 and ta2 ethernet that's very

03:54 important select the right interface for

03:56 the different test agents or they're not

03:58 going to be able to communicate with

03:59 each other very well okay number flows

04:01 we didn't specify that in our criteria

04:03 but we can leave that at one that's fine

04:05 and the codec remember we are selecting

04:07 the gsm efr codec

04:09 and the port let's set that to 1001

04:13 and then the thresholds for aired

04:16 seconds

04:17 remember we want to set that to 5.

04:21 and then the expected dacp value i'm

04:23 going to set that to 10. now recall that

04:26 1-5 is what you can have the mos score

04:29 set to and five being the best so we're

04:32 expecting the best possible results

04:34 so with that i'm going to start the

04:36 monitor

04:38 and we're going to start without a

04:40 report we don't need that here for our

04:41 learning byte if this was a production

04:43 network highly recommend you to set up

04:45 periodic reports

04:48 and the monitor is getting started it

04:50 takes just a minute to get a monitor

04:52 started so it'll just be another few

04:54 seconds and then we'll start seeing some

04:55 results all right so we're seeing some

04:57 results and

04:59 as you can see here we have

05:02 black which means

05:03 50 percent is aired seconds 50 or more

05:06 that is and that's not good that's not

05:10 what we want to see so we have to ask

05:11 ourselves why are we getting this low

05:13 score so recall that with our air

05:15 seconds threshold we have either

05:18 a mos score that is less than five or a

05:21 dacp value that is

05:23 not 10 that is being received and you

05:25 can see here right off the bat that the

05:27 mos score is not five it's at four point

05:29 three which is still really good four is

05:32 good four and above is actually a really

05:33 good mos score but it's not perfect five

05:36 was basically perfect so there's no

05:38 packet loss but the jitter probably

05:40 isn't

05:41 where we want it to be we can select the

05:43 individual streams to get more

05:44 information about this

05:46 you can see here if we remove the air

05:48 second bar the rate we don't need to be

05:50 concerned about

05:53 and

05:54 here you can see the the jitter you can

05:56 see that how that's popping up like that

05:58 that's not what we're liking as far as

06:00 the mos score so that's why that's a

06:01 little lower and if we look at the table

06:03 we see more information here we can see

06:06 the that we have zero loss so that's

06:08 great

06:09 can't have any loss

06:10 we see the jitter here

06:13 and then we have

06:16 so es for dscp how much we're losing any

06:20 at all and we're not losing any so we're

06:21 not having any problems there but we are

06:23 getting 100 aired seconds and that's

06:25 because that mos value so let's go ahead

06:27 and close that and then let's

06:30 edit this on the fly that's the great

06:31 thing about monitors you can just edit

06:33 on the fly and let's change this to four

06:35 instead of five so recall that four is

06:38 really good

06:41 and let's save that and that bumps us

06:43 right back out to the monitor and this

06:45 will take just a minute to update and

06:47 you can see now we're getting some more

06:49 results we can see that the aired

06:51 seconds is no longer black

06:54 that is the air seconds bar we're

06:55 getting green here now on both streams

06:57 and that's really good to see and so we

06:59 can select this here

07:01 and let's go to table or we go back to

07:03 graph you can see here where the error

07:04 second started

07:06 not being

07:07 100 percent

07:09 in our case and we can look at the

07:12 error seconds here as

07:14 well you see it's 100

07:17 we can scroll down now we see it dropped

07:18 down to 60. and if we close that

07:21 and then open it again

07:23 that should actually look a little

07:24 better

07:25 we should have some additional results

07:27 in there

07:29 oh actually i misspoke we're only

07:30 showing errors so let's click show all

07:33 we should get a little better

07:34 information in there so there let me

07:36 scroll down

07:42 and then let's scroll over

07:44 and there you go so we can see where we

07:46 had sixty percent that's where things

07:48 are getting back to better then we have

07:49 zero percent zero percent zero percent

07:51 zero percent for that ten seconds there

07:53 that resolution window we had sixty

07:54 percent error seconds and then we

07:56 dropped to zero and the reason behind

07:58 that we have the mos score set to four

08:00 instead of five as far as what we want

08:03 for the air second threshold so things

08:05 look good there

08:09 so that does bring us to the end of this

08:10 learning byte so this learning byte we

08:12 demonstrated how to configure and run a

08:14 voip udp monitor so as always thanks for

08:18 watching

08:21 visit the juniper education services

08:23 website to learn more about courses

08:26 view our full range of classroom online

08:29 and e-learning courses

08:31 learning paths

08:33 industry segment and technology specific

08:36 training paths

08:37 juniper networks certification program

08:40 the ultimate demonstration of your

08:42 competence and the training community

08:45 from forums to social media join the

08:47 discussion

08:54 you