Paragon Active Assurance – Multicast UDP Monitor

0:00 [Music]

0:12 hello my name is zach gibbs and i'm a

0:15 content developer within education

0:17 services inside juniper networks and

0:20 today we will be going through the

0:22 paragon active assurance multicast udp

0:25 monitor learning byte

0:28 all right so here is the topology for

0:30 our learning byte and there are three

0:32 main devices we want to

0:34 talk about we have ta1 ta2 and the

0:37 control center so ta1 and ta2 use e0 to

0:41 connect to the management network which

0:43 then provides connectivity with the

0:45 control center

0:46 and then ta1 and ta2 use eth1 to connect

0:50 into the service provider network

0:52 all right so with that what are we going

0:54 to do we need to configure a multicast

0:56 udp task and so first you probably know

0:59 what multicast traffic is but just in

1:01 case you don't multicast is normally

1:03 used as a transport mechanism for

1:05 services such as iptv or even updating

1:08 pcs at the same time

1:10 so by using the multicast udp task

1:13 you're able to figure out

1:14 how your network handles multicast

1:16 traffic with things like end-to-end

1:18 delay jitter and packet loss

1:21 and so this task multicast udp task

1:24 sends udp multicast traffic from a

1:27 server which in our case is going to be

1:29 ta1 and then ta2 acts as a client to

1:33 receive the multicast traffic

1:35 and so what happens when the test starts

1:37 well first ta1 which as we know is the

1:39 server generates multicast traffic

1:41 towards the destination multicast

1:43 address

1:44 which in our case is 239.1.1.100

1:48 then ta2 which is the client tries to

1:52 join the multicast address and if ta2 is

1:54 successful at joining that multicast

1:56 address things like one way delay jitter

1:59 packet loss and packet miss orderings

2:01 that can be recorded now one thing i do

2:03 want to point out here is this would

2:05 make a lot more sense if we had more

2:06 than one client because we are talking

2:07 multicast what is the point of multicast

2:10 if we only have one client right well

2:12 the concept still holds true with just

2:14 one client ta2 is our one client ta1 is

2:16 the server and i don't have another test

2:19 station in the test bed right now so

2:20 that's why i'm using just one client

2:22 and so with that it'll still prove the

2:24 concept with only having one client but

2:27 typically you'd have multiple clients at

2:29 least two probably 10 maybe 100 clients

2:32 that would be subscribing to that

2:33 multicast address so just keep that in

2:35 mind okay so with that what are some

2:37 things we haven't talked about duration

2:38 we're going to set to 30 seconds the

2:39 server and the client we talked about

2:41 the rate we're going to set at 10

2:42 megabits per second the multicast

2:44 address of

2:45 239.1.1.100 we did talk about the server

2:48 port it's going to be three one two four

2:50 and then for error seconds threshold

2:52 we're just going to be using zero

2:53 percent loss we could also say something

2:55 about the jitter or the delay or

2:57 whatever in the air second thresholds

2:59 but with this learning about we're going

3:00 to keep it simple and put it at zero

3:02 percent loss for the only error second

3:03 threshold metric and then we'll run the

3:05 monitor and examine the results and so

3:08 with that let's go ahead and jump to the

3:10 control center web interface and get

3:12 this going

3:14 all right so here is the control center

3:16 web interface let's go to

3:18 monitoring and then new monitor

3:21 and we'll name this

3:23 multicast udp lb lb short for learning

3:26 byte

3:27 and then we'll select the multicast udp

3:30 task and you can see here that this is a

3:32 puzzle piece task which means we can add

3:34 additional parallel tasks and we could

3:36 run

3:37 other tasks at the same time as this

3:39 task but for this learning byte we are

3:41 focusing on the multicast udp task

3:43 okay so with that we need to fill out

3:45 the general section we need to specify

3:46 the server which is going to be ta1

3:48 we're going to be using the eth1

3:49 interface for this we want to make sure

3:51 we select the right interface if we were

3:53 to select each zero that would be the

3:55 management network and we don't want to

3:57 do that

3:58 okay so then the clients this is just

4:00 going to be ta2 the eth1 interface and

4:03 then the rate we're going to set that at

4:05 10 megabits per second multicast address

4:08 let's say 239 1.1.100

4:11 and the port recall that is three

4:13 one two four

4:15 and the loss is already defaulted to for

4:17 errored second threshold that is is

4:19 already defaulted to zero percent again

4:21 we could set the delay and the jitter

4:24 for air second threshold and also the

4:26 expected dscp we make sure that we're

4:28 receiving expected dcp value

4:31 but we're not going to worry about that

4:32 right now

4:33 and so with that let's go ahead and

4:34 start the monitor

4:36 and we don't need to start this with a

4:37 report if you are running a monitor in a

4:40 production environment highly recommend

4:41 it but for our learning bite it's not

4:43 necessary

4:45 and with any monitor there's a bit of a

4:47 setup time it is a little longer than a

4:49 test so this just takes a little bit of

4:51 extra time to get the setup going and

4:52 whatnot you can see here it's beginning

4:54 to start just be a few more seconds

4:57 okay you can see the monitor has started

4:59 and things are looking good we have

5:00 complete green no aired seconds so we

5:02 are getting no loss because recall that

5:03 we were only using the loss percentage

5:05 for the error second thresholds values

5:08 and so things are looking good we got

5:09 green across the board light green

5:11 across the board i guess at a start

5:12 green we're getting a little bit of a

5:14 loss a little bit of air settings but

5:15 that's not happening so you can see

5:17 things look good here and we can click

5:18 on the individual stream now one thing i

5:20 want to point out here this might seem a

5:21 little strange because it says ta1 eth1

5:24 ipv4 then giving us an arrow to ta2 eth1

5:28 ipv4 and that might seem a little

5:30 strange because that almost sounds like

5:32 it's unicast traffic but it's not what's

5:34 happening is ta1 is sending

5:36 traffic to that multicast address and

5:38 then ta2 is subscribing to that

5:40 multicast address and receiving that

5:41 traffic so with that let's go ahead and

5:43 click on the stream we can look at some

5:45 more information and we can see here

5:47 that we have some information in the

5:49 graph we can see here that we have the

5:51 air second bar the rate and the average

5:53 delay in milliseconds selected and you

5:55 can see here that at the top we can see

5:57 the rate

5:58 and

5:59 which kind of makes sense the average

6:00 delay is so low that nothing you can't

6:02 really see anything so let's go ahead

6:04 and remove that rate we know the rate is

6:05 good we're not getting any loss so let's

6:07 remove the rate from the graph

6:09 and okay great now we're seeing the es

6:11 bar down here uh there's no air second

6:14 so it's all green that's great and then

6:16 we're seeing the average delay in

6:17 milliseconds you can actually see what's

6:18 going on there

6:19 that's pretty low that's that's good and

6:22 then we could also select the max delay

6:24 and the min delay and you can actually

6:26 see that showing up there you can see

6:27 how that's

6:29 popping up here

6:30 you can see the max delay is kind of up

6:32 higher there but the average is pretty

6:34 good

6:35 let's deselect those and then we can

6:36 select the jitter as well to get some

6:38 information on that so you can see how

6:40 the jitter is coming through and then we

6:41 can look at the table the table will

6:43 actually show us that information as

6:44 well now it doesn't show anything

6:46 because there's no errors so we need to

6:47 show the

6:48 or use the show all command and we're

6:50 going to have to scroll down a bit to

6:51 see it because remember it takes a while

6:53 to get it going

6:54 and so with the monitor that is so we do

6:56 have to scroll down to start seeing the

6:58 results so you can see the results here

7:00 you can see the rate the loss miss order

7:02 packets none delay average delay delay

7:05 max things like that so you can see the

7:07 individual statistics that are recorded

7:09 for this monitor and remember since this

7:11 is a monitor it has a 10 second

7:13 resolution and so it's recording the

7:15 statistics every 10 seconds

7:18 so this does bring us to the end of this

7:20 learning by in this learning but we

7:21 demonstrated how to configure and run a

7:23 multicast udp monitor so as always

7:26 thanks for watching

7:29 visit the juniper education services

7:32 website to learn more about courses

7:34 view our full range of classroom online

7:38 and e-learning courses

7:40 learning paths

7:41 industry segment and technology specific

7:44 training paths

7:46 juniper networks certification program

7:48 the ultimate demonstration of your

7:50 competence and the training community

7:53 from forums to social media join the

7:56 discussion