All language subtitles for Free-CCNA-QoS-Part-2-Day-47-CCNA-200-301-Complete-Course_en

Welcome to Jeremy’s IT Lab. This is a free,\n

these videos, please subscribe to follow along\n

a comment, and share the video to help spread\n

In this video we will continue our study of\n

voice VLANs and Power over Ethernet in the\n

QoS in the second half. Hopefully you understand\n

certain traffic such as voice and video traffic\n

also to ensure it gets the required bandwidth.\n

terms in exam topic 4.7, such as classification,\n

shaping. Note that you don’t need to know\n

As long as you understand the topics I introduce\n

Here’s what we’ll cover in today’s video.\n

how routers and switches identify which traffic\n

cover queuing and congestion management. I\n

but let’s go more in depth. Finally shaping\n

rate at which traffic enters or exits an interface.\n

practice question from Boson Software’s\n

Let’s begin with classification. As you\n

kinds of network traffic priority over others\n

organizes network traffic, meaning packets,\n

is that important for QoS? Classification\n

certain types of traffic, you have to identify\n

So, how can we classify traffic? There are\n

are some examples. You can use an ACL. Traffic\n

treatment, for example it could be treated\n

is denied by the ACL will not be given that\n

dynamic NAT. The ACL isn’t being used to\n

being used to identify certain traffic. There\n

Sometimes just looking at the Lower-layer\n

and IP address, isn’t enough to identify\n

what’s called a ‘deep packet inspection’\n

all the way up to Layer 7 to identify the\n

focus on methods like these in this video.\n

specific fields used for this purpose, for\n

them briefly in previous videos, but finally\n

code point, field of the 802.1Q tag can be\n

Keep in mind that this field can only be used\n

tag is added to the Ethernet header. Then\n

Code Point, field of the IP header. It can\n

traffic. Okay, let’s take a look at each\nof these.

First up, PCP. Here you can see an Ethernet\n

field is in this dot1q tag. Here are the fields\n

in the course, and the PCP field is this 3-bit\n

this field referred to as CoS, class of service.\n

of QoS, CoS just refers to this part of the\n

So, let’s see how it has been defined. Well,\n

0 to 7. This is how they are defined. You\n

but I recommend remembering 0 for best effort,\n

and 5 for voice. Regarding 0, best effort,\n

means there is no guarantee that data is delivered\n

regular traffic, not high-priority. This is\n

have a value of ‘0’ in the PCP field.\n

in the previous video, let me point out something\n

traffic as PCP3. Call signaling traffic is\n

the call is established, the actual voice\n

as PCP5. I put the term mark in bold because\n

to mark traffic is to set the value in the\n

at those markings and use them to classify\n

etc. So, when an IP phone marks its voice\n

routers and switches to classify those packets\nas high-priority.

Here’s a simple network with a couple PCs\n

to demonstrate a very important point about\n

PCP field is found in the dot1q header, it\n

types. First, the obvious one is trunk links.\n

dot1q, unless the traffic is in the native\n

being used. However, as I explained in the\n

well even over access links. So, assuming\n

and phones can communicate with each other\n

of these links are either trunk links or access\n

between the phones and the switches are access\n

the phones will be tagged. And these two connections\n

10 or 20. So, over these connections devices\n

to tell other devices to treat the traffic\n

receiving that marked traffic can classify\n

priority based on the PCP marking. Let me\n

traffic from the PCs is not tagged with dot1q\n

this link. In addition to that, all traffic\n

destinations will not have a dot1q tag. So,\n

with a PCP value, PCP cannot be used to classify\n

using PCP, a limitation which the Layer 3\n

So now let’s look at how marking and classification\n

is a byte that is referred to as the ToS byte,\n

the traffic class byte used for QoS, but let’s\n

here, the second one in the header after the\n

the ToS byte consists of two fields, DSCP,\n

explicit congestion notification. However,\n

Here’s the old use of the ToS byte. Three\n

field. It was used to mark packets according\n

dot1q header. The remaining 5 bits were mostly\n

is that they weren’t really used. So, 3\n

just like in the PCP field of the dot1q header.\n

the one shown above, is this. 6 bits for DSCP\n

total of 64 values, which gives a lot of flexibility\n

Before going in depth about DSCP, let’s\n

markings are similar to PCP. 6 and 7 are reserved\n

traffic. That refers to traffic like OSPF\n

voice traffic is marked as IPP 5, interactive\n

used for voice signaling traffic, for setting\n

is used for best effort traffic, regular data\n

So, with 6 and 7 reserved, only 6 possible\n

many networks, the QoS requirements of some\n

And because IPP only used 3 bits of the ToS\n

it was decided that an extra 3 bits would\n

Now let’s take a look at DSCP. RFC 2474,\n

field, and then other ‘DiffServ’, differentiated\n

of the field. With IPP updated to DSCP, new\n

Why is that? By having generally agreed upon\n

QoS design and implementation is simplified,\n

because they agree upon the markings that\n

too. So, a few sets of industry-standard markings\n

me say that it might feel a bit overwhelming\n

flashcards to help you memorize them, but\n

are some of the more common ones. I’ll point\n

of the markings. So, you should be aware of\n

forwarding, DF. This is the marking for best\n

QoS requirements. Then there is EF, Expedited\n

low loss, latency, and jitter, which is usually\n

AF. This isn’t one marking, but a set of\n

selector, CS, which is a set of 8 standard\n

We won’t cover QoS configuration because\n

you this. I configured a ‘class-map’ called\n

used to identify which traffic you want to\n

that I want to match traffic based on the\n

out the options. At the top is the choice\n

from 0 to 63. Then below that are the 12 AF\n

for example AF11 has a binary value of 001\n

the binary values are displayed on the right.\n

shown here. That’s because the other one,\n

000 000. Finally at the bottom is EF, expedited\n

First up, DF and EF. DF is used for best-effort\n

So, all 6 of these bits will be set to 0.\n

continue, I’ll write the decimal value of\n

traffic like sending an email or downloading\n

DSCP field of the IP header, indicating that\n

Then there is EF, expedited forwarding. EF\n

latency, and jitter. Typically, voice traffic\n

46. So, this is how it looks in binary, 1\n

to remember. DF is used for best-effort traffic,\n

requiring low-loss, low-latency, and low-jitter,\n

is 46. Now things will get a bit more complicated.

So we just looked at default, also called\n

forwarding. AF can be a little tricky to understand,\n

AF defines four traffic classes. All packets\n

class number means higher priority, their\n

than lower-priority packets. Then, within\n

precedence. A higher drop precedence means\n

due to WRED. Now, notice that I’ve set up\n

up, in AF this bit on the end is always set\n

this, it might just be because the designers\n

These two bits in red represent the drop precedence.\n

Now, when we write an AF value, it’s written\n

of the class and Y being the decimal number\n

at an example to see how to do it.

So, we have a binary DSCP value of 001 010.\n

AF X Y, with X being the decimal number of\n

the drop precedence. To do that, we split\n

two bits for the drop precedence. So, the\n

of the drop precedence is also 1. So, this\n

you like. Now, really this is just a 6-bit\n

values up top, 1 2 4 8 16 and 32, we can calculate\n

value as DSCP10. AF is just a set of standard\n

than simply having 64 DSCP values with no\n

Let’s do some more practice. Now we have\n

time it’s class 1, drop precedence 2. So\n

this as a normal decimal DSCP value? Here\n

case, the DSCP value is also written as 12.\n

Here’s another one. Try to figure it out\n

how could we write it as a normal decimal\n

showing you the value of each binary bit.\n

each bit. This is AF2 3, or AF23. Written\n

you were able to figure that out. If not that’s\n

This time we have a binary DSCP value of 011\n

written as a normal DSCP value? Let’s check.\n

is AF3 2, or AF32. Written as a normal DSCP\n

Okay, here’s one more for practice. Again,\n

it as a normal DSCP value? Let’s check.\n

3, so this is AF4 3, or AF43. Written as a\n

is equivalent to DSCP 38. By the way, within\n

5, 6, or 7. Okay, so you should be able to\n

values. If you want a quick way to calculate\n

here’s the formula. 8X plus 2Y, again the\n

reason for this is that 8 is the lowest value\n

value of the ‘Y’ portion. So to calculate\n

understand the binary underneath it all, whether\n

IPv6, matching with ACLs, QoS, whatever. But\n

nice to have some shortcuts like this, 8X\nplus 2Y.

So, here’s a summary of all of the AF values.\n

So, within these AF values traffic marked\n

the highest priority class, but has the lowest\n

marked as AF13 gets the worst treatment, being\n

drop precedence. I will include flashcards\n

them from the AF values to the regular DSCP\n

them, but you should be able to calculate\n

Okay, that’s all for AF. I hope my explanation\n

Finally, let’s look at CS. Fortunately,\n

CS, class selector, defines eight DSCP values\n

that backward compatibility work? The three\n

and the original IPP bits are used to make\n

again. Notice that the three bits on the right\n

the original IP precedence bits, can be either\n

three bits, 1 2 and 4. So, IPP gives us values\n

CS are CS0, CS1, CS2, 3, 4, 5, 6, and 7. Very\n

DSCP value of these. Although CS is a way\n

8 values, they are still DSCP values. Fortunately\n

just 8 multiplied by the CS number. CS0 is\n

32, CS5 is 40, CS6 is 48, and CS7 is 56. Okay,\n

Okay, so we’ve got DF, EF, AF, and CS. How\n

use them? RFC 4954 was developed with the\n

together and standardize their use. Many specific\n

here are a few key ones. Voice traffic is\n

delay, jitter, and loss. Interactive video\n

or 3. Streaming video is marked as AF3x, high\n

is marked as DF, a DSCP value of 0. There\n

the RFC, do a google search for it if you’re\n

to the engineer designing the QoS policy of\n

These are just standard recommendations.

Okay, let’s move on to take a quick look\n

boundary of a network defines where devices\n

received messages. If the markings are trusted,\n

without changing the markings. But if the\n

change the markings according to the configured\n

boundary is here, at SW1. Phone1 sends a message\n

is referring to the PCP field in the dot1q\n

you should be familiar with both terms. Anyway,\n

because it’s from outside of the trust boundary.\n

DF and the CoS marking to 0, before forwarding\n

just the DF marking because there is no dot1q\n

Usually it’s best to trust the markings\n

If an IP phone is connected to the switchport,\n

to the IP phone. This is done via configuration\n

by the way, not directly on the phone itself.\n

PC’s traffic with a high-priority marking\n

changed according to the configured policy.\n

be trusted by the switch. In this case, if\n

SW1 will trust those markings and not change\n

sends an EF-marked packet, the switch should\n

the EF marking, DSCP 46, to DF, DSCP 0. We\n

treated with the same priority as the voice\n

for trust boundaries. You don’t need to\n

just be aware of the concept of trust boundaries\n

Okay, that’s all for classification and\n

congestion management. I already introduced\n

bit more to be covered for the CCNA. For review,\n

a faster rate than it can forward the traffic\n

are placed in that interface’s queue as\n

becomes full, packets that don’t fit in\n

tail drop. RED and WRED, which I already introduced,\n

the image I showed last video. The router\n

the queue gets full and packets start getting\n

However, an essential part of QoS is the use\n

And this is where classification really plays\n

on various factors, for the example the DSCP\n

things, and then place the traffic in the\n

only able to forward one frame out of an interface\n

which queue traffic is forwarded from next.\n

certain queues more priority than others.\n

becomes clear. Here’s that same example\n

router’s interface forwarding the traffic\n

Here’s how it works. Ingress traffic is\n

incoming traffic, traffic entering the router.\n

which interface to send it out of, as well\n

it classifies the traffic and places it into\n

are four queues and traffic is classified\n

the DSCP marking. Then the scheduler decides\n

in which order, and the router forwards the\n

oversimplification, but its basically how\n

to forward the packet out of, it is classified,\n

A common scheduling method is weighted round-robin.\n

each queue in order, cyclically. And weighted\n

queues each time the scheduler reaches that\n

Next, here’s a term you definitely should\n

is a popular method of scheduling, using a\n

each queue a certain percentage of the interface’s\n

let’s put these together. Here’s the process\n

schedule it, and transmit. The device is using\n

a certain amount of traffic from each queue\n

a guaranteed minimum amount of bandwidth,\n

is getting a lot more advanced than just a\n

ideal. Specifically it’s not ideal for voice\n

a guaranteed minimum amount of bandwidth,\n

even the high-priority voice and video queues\n

To solve that, we can configure LLQ, low latency\n

as strict priority queues. Strict priority\n

the scheduler will always take the next packet\n

very effective for reducing the delay and\n

as traffic enters the priority queue, the\n

that same diagram, but this time the top queue\n

is traffic in the queue, so the scheduler\n

the weighted round-robin scheduling of the\n

here. LLQ has the downside of potentially\n

in the designated strict priority queue. The\n

traffic. Policing, which I will cover in the\n

allowed in the strict priority queue so that\n

Okay, so in this section we expanded on the\n

video and examined the use of multiple queues.\n

looking at the DSCP value, then places it\n

example using weighted round-robin logic,\n

With the addition of LLQ, a strict priority\n

priority packets. And by the way, within each\n

like RED or WRED can be used to avoid tail\n

Okay, here are the final topics for today,\n

basically just have to understand what these\n

them in one slide. Traffic shaping and policing\n

In the previous examples of queuing and scheduling\n

at full capacity, or beyond full capacity\n

there are situations in which it is desirable\n

actual maximum capacity of the link. Shaping\n

rate goes over the configured rate. So, this\n

instead of the actual capacity of the link\n

traffic rate configured on the link. Policing\n

if the traffic rate goes over the configured\n

traffic over the configured rate is allowed\n

data applications which are ‘bursty’ in\n

they tend to send data in bursts. Just like\n

traffic allowed is also configurable. And\n

to allow for different rates for different\n

to limit the rate traffic is sent or received?\n

sample network. A customer router is connected\n

The customer configures shaping outbound on\n

ISP configures policing inbound on the G0/0\n

of a reason why this might be done? Although\n

1000 megabits per second, perhaps this customer\n

So the ISP says, you paid for a 300 megabit\n

incoming traffic to 300 megabits per second.\n

faster than 300 megabits per second it will\n

outgoing traffic to 300 megabits per second.\n

and policing, but this is a common use of\nthese tools.

Okay, that was a lot of material to cover.\n

on to the quiz. First, we covered classification\n

different kinds of traffic so that you can\n

priority. Marking refers to setting the values\n

3 headers for use in classification. We covered\n

byte in the IP header, including IP Precedence\n

also introduced the concept of trust boundaries.\n

management, which I introduced in the last\n

queues, weighted round-robin scheduling, CBWFQ,\n

doesn’t actually do anything on its own.\n

LLQ to make the devices treat those packets\n

and policing, which are both tools to control\n

until the end of the quiz for a bonus practice\n

CCNA. Okay, let’s go to quiz question 1.

SLIDE30\nWhich of the following CoS markings are consistent

with standard practice? Select three. Okay,\n

The answers are B, CoS 0 for best effort.\n

Here’s that chart again showing the PCP\n

values, and their traffic types. In your networks\n

but this is standard practice. Okay, let’s\ngo to question 2.

SLIDE31\nWhat bit pattern would you find in the DSCP

field of a packet marked as EF? Pause the\n

Okay, the answer is D, 101 110. Here it is,\n

which is used for traffic requiring low delay,\n

of 46, so the bit pattern is 101 110. Okay,\n

SLIDE32\nWhich of the following AF markings provides

the best service? Pause the video now to think\nabout the answer.

The answer is B, AF41. Here is the table of\n

priority class and it has the lowest drop\n

queue, but it has a higher drop precedence.\n

only uses classes 1, 2, 3, and 4. Okay, let’s\n

SLIDE33\nWhich of the following statements represents

general best practice regarding QoS? Pause\n

Okay, the answer is A, trust markings from\n

IP phones will typically mark their voice\n

should be trusted because voice traffic requires\n

PCs should not be trusted, though. Traffic\n

as low priority so that it doesn’t fill\n

traffic. Now, apps like Zoom or WebEx used\n

we can mark those packets at the switch or\n

SLIDE34\nWhich of the following creates a strict priority

queue for data that requires low delay, jitter,\n

there are packets in that queue the scheduler\n

in the other queues. Okay, that’s all for\n

question in Boson Software’s ExSim for CCNA.\n

There are supplementary materials for this

video. There is a flashcard deck to use with\n

a packet tracer practice lab so you can get\n

isn’t actually part of the CCNA exam and\n

think it will be beneficial to see how it’s\n

will be in the next video. Sign up for my\n

and I’ll send you all of the flashcards\n

SLIDE36\nBefore finishing today’s video I want to

thank my JCNP-level channel members. To join,\n

video. Thank you to Justin, Christopher, Sam,\n

Serge, Njoku, Viktor, Roger, Raj, Kenneth,\n

Gustavo, Benjamin, Prakaash, Nasir, Erlison,\n

Mark, Yousif, Boson Software, Devin , Yonatan,\n

incorrectly, but thank you so much for your\n

at the time of recording by the way, April\n

name isn’t on here don’t worry, you’ll\nbe in future videos.

Thank you for watching. Please subscribe to\n

and share the video with anyone else studying\n

check the links in the description. I'm also\n

or Basic Attention Token, tips via the Brave\n