Tag: Cisco

In VLAN configurations, trunking allows multiple VLANs to be transmitted over a single physical link between switches. Trunking can be set up using static trunking or dynamic trunking.

---

1. Static Trunking

Static trunking means manually configuring a switch port as a trunk. This ensures that the port remains in trunk mode, allowing multiple VLANs to pass through it.

Advantages of Static Trunking

✅ More secure (prevents unauthorized devices from negotiating trunks).
✅ No risk of dynamic negotiation failures.
✅ Preferred in enterprise networks for stability.

Configuration of Static Trunking (Cisco Switch Example)

bashCopyEditinterface GigabitEthernet0/1
  switchport mode trunk
  switchport trunk allowed vlan 10,20,30
  switchport trunk native vlan 99
  exit

👉 This command sets GigabitEthernet0/1 as a static trunk, allowing VLANs 10, 20, and 30, and setting VLAN 99 as the native VLAN.

---

2. Dynamic Trunking (DTP – Dynamic Trunking Protocol)

Dynamic trunking allows ports to negotiate whether to become a trunk or remain an access port. Cisco switches use DTP (Dynamic Trunking Protocol) to automate trunk formation.

DTP Modes

Mode	Behavior
Access	Forces the port into access mode (no trunking).
Trunk	Forces the port into trunk mode (like static trunking).
Dynamic Auto	Waits for the other side to initiate trunking but does not actively negotiate.
Dynamic Desirable	Actively tries to negotiate trunking with the other device.

Configuration of Dynamic Trunking (Cisco Example)

interface GigabitEthernet0/2
  switchport mode dynamic desirable
  exit

👉 The desirable mode will actively try to establish a trunk if the other side supports it.

interface GigabitEthernet0/3
  switchport mode dynamic auto
  exit

👉 The auto mode waits for the other side to initiate trunking. If both sides are set to auto, the trunk will not form.

---

3. Key Differences: Static vs. Dynamic Trunking

Feature	Static Trunking	Dynamic Trunking (DTP)
Configuration	Manually set to trunk mode	Uses DTP to negotiate trunking
Security	More secure	Less secure (vulnerable to VLAN hopping attacks)
Stability	Always trunk	May fail to negotiate a trunk
Best Use Cases	Enterprise networks, high-security environments	Simple setups or mixed vendor environments

Best Practice Recommendation

🔹 Disable DTP on all trunk ports and configure static trunking for security.
🔹 Use switchport nonegotiate to prevent DTP from running:

interface GigabitEthernet0/4
  switchport mode trunk
  switchport nonegotiate
  exit

1. What is 802.1Q?

IEEE 802.1Q is the standard for VLAN tagging in Ethernet networks. It allows multiple VLANs to exist on a single physical network by inserting a VLAN tag in the Ethernet frame header. This tagging enables switches to distinguish between VLANs and forward traffic accordingly.

Key Features of 802.1Q:

• Adds a 4-byte VLAN tag to Ethernet frames.
• Supports up to 4094 VLANs (VLAN IDs 1-4094).
• Trunk links carry multiple VLANs between switches.
• Defines a Native VLAN (untagged traffic).

---

2. What is a Native VLAN?

A Native VLAN is the VLAN that carries untagged traffic on a trunk port. Any frame received without a VLAN tag is assumed to belong to the Native VLAN of that trunk.

By default, VLAN 1 is the Native VLAN on most switches, but best practice is to change it to another VLAN for security reasons.

Example Configuration (Changing the Native VLAN):

bashCopyEditinterface GigabitEthernet0/1
  switchport mode trunk
  switchport trunk native vlan 99
  exit

👉 This sets VLAN 99 as the Native VLAN for trunk port GigabitEthernet0/1.

---

3. What is an Allowed VLAN?

An Allowed VLAN is a VLAN that is explicitly permitted on a trunk link. By default, all VLANs are allowed on a trunk, but administrators can restrict the VLANs allowed on a specific trunk port to enhance security and reduce unnecessary traffic.

Example Configuration (Restricting Allowed VLANs on a Trunk Port):

bashCopyEditinterface GigabitEthernet0/2
  switchport mode trunk
  switchport trunk allowed vlan 10,20,30
  exit

👉 This command allows only VLANs 10, 20, and 30 on trunk port GigabitEthernet0/2.

---

Summary of Key Concepts

Concept	Description
802.1Q	Standard for VLAN tagging on Ethernet frames.
Native VLAN	The VLAN for untagged traffic on a trunk port (default is VLAN 1).
Allowed VLANs	VLANs that are explicitly permitted on a trunk link.

A VLAN (Virtual Local Area Network) is a logical segmentation of a physical network that allows multiple networks to exist within the same physical infrastructure. VLANs enhance network performance, security, and manageability by grouping devices logically rather than based on physical location.

Benefits of VLANs

How to Configure VLANs

Step 1: Define VLANs on a Switch

Access the switch using the CLI (Command Line Interface) or Web GUI.

Enter privileged EXEC mode:enable

Enter global configuration mode:configure terminal

Create a VLAN and assign it a number:

vlan 10
name Sales
exit

for additional VLANs as needed.

Step 2: Assign Ports to VLANs

Enter interface configuration mode for a specific port:

interface GigabitEthernet0/1

Assign the port to VLAN 10 (Access Mode):

switchport mode access switchport

access vlan 10

exit

Repeat for other interfaces and VLANs as needed.

Step 3: Configure Trunk Ports (for VLAN Communication Between Switches)

Enter interface configuration mode for the uplink port:

Edit interface GigabitEthernet0/24

Set the port as a trunk:

switchport mode trunk

switchport trunk allowed vlan 10,20,30

exit

Step 4: Verify VLAN Configuration

Check VLAN assignments:

show vlan brief

Check trunk status:

show interfaces trunk

Verify VLAN connectivity using ping or other network tools.

1. Introduction

Weighted Random Early Detection (WRED) is a congestion avoidance QoS mechanism that randomly drops packets before the queue is full, helping to prevent global TCP synchronization and optimize network performance.

📌 Key Benefits of WRED:
✅ Prevents queue tail drops, reducing congestion
✅ Improves TCP traffic performance
✅ Differentiates traffic by priority levels
✅ Works best in TCP-based networks

---

2. How WRED Works

WRED monitors queue depth and randomly drops packets based on:

• Queue occupancy (how full the queue is)
• Packet marking (DSCP/IP Precedence)
• Minimum and maximum drop thresholds

🔹 WRED Drop Logic:

Queue Status	Behavior
Queue < Min Threshold	No drops ✅
Queue between Min & Max	Random drops based on priority 🎲
Queue > Max Threshold	All new packets dropped (Tail Drop) 🚫

📌 Higher-priority traffic has higher thresholds, reducing drop probability.

---

3. WRED vs. Tail Drop

Feature	WRED	Tail Drop
Packet Drop Behavior	Gradual, based on queue depth	Drops all packets when full
Effect on TCP	Prevents global TCP synchronization	Causes synchronization issues
QoS Differentiation	Prioritizes important traffic	No differentiation

---

4. Configuring WRED on a Cisco Router

A. Basic WRED Configuration

To enable WRED on an interface:

interface GigabitEthernet1/0/1
 random-detect

📌 This applies default WRED settings, treating all traffic equally.

---

B. WRED with IP Precedence-Based Drop Probabilities

To prioritize higher-priority traffic:

interface GigabitEthernet1/0/1
 random-detect precedence 3 20 40
 random-detect precedence 5 30 60

📌 Explanation:

• Precedence 3 packets → Start dropping at 20% queue, full drop at 40%.
• Precedence 5 packets → Start dropping at 30%, full drop at 60%.
• Higher precedence = Less chance of being dropped.

---

C. WRED with DSCP-Based Drop Probabilities

To configure WRED based on DSCP values:

interface GigabitEthernet1/0/1
 random-detect dscp-based

Then, define DSCP drop thresholds:

• DSCP 10 (low priority) → Drops start at 20%, full drop at 50% queue depth.
• DSCP 46 (VoIP, high priority) → Drops start at 40%, full drop at 70%.

---

D. WRED with Class-Based QoS

1️⃣ Define a Traffic Class:

class-map MATCH-VIDEO
 match ip dscp af41

2️⃣ Create a Policy with WRED:

policy-map WRED-POLICY
 class MATCH-VIDEO
  random-detect dscp-based

3️⃣ Apply Policy to an Interface:

interface GigabitEthernet1/0/1
 service-policy output WRED-POLICY

📌 Now, WRED is applied only to DSCP AF41 traffic!

---

5. Verifying WRED

✅ Check WRED settings on an interface:

show interfaces GigabitEthernet1/0/1 random-detect

✅ Monitor packet drops with WRED:

show policy-map interface GigabitEthernet1/0/1

---

6. Summary

Scenario	Configuration
Enable WRED on an interface	random-detect
Prioritize traffic based on IP Precedence	random-detect precedence X min max
Prioritize traffic based on DSCP	random-detect dscp-based
Class-Based WRED	policy-map WRED-POLICY + random-detect dscp-based
Verify WRED settings	show interfaces random-detect

🚀 WRED improves congestion management by preventing queue overflows and TCP synchronization issues!

1. Introduction

Traffic policing is a QoS mechanism that limits the rate of traffic by dropping or marking excess packets when the defined limit is exceeded. Unlike traffic shaping, which buffers excess traffic, policing discards or reclassifies packets immediately.

📌 Key Benefits of Traffic Policing:
✅ Enforces bandwidth limits on applications/users
✅ Prevents network abuse (e.g., users setting high DSCP values)
✅ Protects critical traffic by limiting non-essential traffic
✅ Can mark or drop excess traffic to maintain QoS policies

---

2. How Does Policing Work?

Policing uses the Token Bucket Algorithm to monitor the rate of incoming traffic.

📌 Behavior of Policing:

• Traffic within the limit → Allowed ✅
• Traffic exceeding the limit → Dropped or marked 🚫
• No buffering (unlike shaping)

Policing vs. Shaping

Feature	Policing	Shaping
Purpose	Limits and enforces a strict rate	Smooths out bursts
Effect on Excess Traffic	Drops or marks packets	Buffers packets in a queue
Delay Impact	No delay	Can introduce delay
Best Use Case	Ingress (incoming) traffic	Egress (outgoing) traffic

---

3. Configuring QoS Policing on Cisco Routers & Switches

Cisco uses Class-Based Policing to enforce bandwidth limits.

A. Policing All Traffic on an Interface

To police all traffic on an interface to 5 Mbps:

interface GigabitEthernet1/0/1
police 5000000 conform-action transmit exceed-action drop

• conform-action transmit → Allowed traffic goes through.
• exceed-action drop → Traffic beyond 5 Mbps is dropped.

---

B. Class-Based QoS Policing

1️⃣ Define a Class to Match Traffic:

class-map MATCH-VIDEO
 match ip dscp af41

2️⃣ Create a Policy Map to Apply Policing:

policy-map POLICE-VIDEO
 class MATCH-VIDEO
  police 2000000 250000 exceed-action drop

3️⃣ Apply to an Interface:

interface GigabitEthernet1/0/1
 service-policy input POLICE-VIDEO

🔹 Explanation:

• Limits AF41-marked traffic (video) to 2 Mbps.
• Bursts up to 250 Kbps are allowed.
• Excess traffic is dropped.

---

C. Marking Instead of Dropping (Two-Color Policing)

Instead of dropping excess traffic, mark it with a lower priority (DSCP 10):

policy-map POLICE-WEB
 class MATCH-WEB
  police 1000000 200000
   conform-action transmit
   exceed-action set-dscp-transmit 10

📌 This means:

• Traffic under 1 Mbps is transmitted normally.
• Traffic above 1 Mbps is marked as DSCP 10 (lower priority) instead of being dropped.

---

D. Three-Color Policing (CIR, PIR, Drop)

Three-color policing allows three actions:
✔ Conform (Transmit ✅)
✔ Exceed (Mark lower priority ✏️)
✔ Violate (Drop 🚫)

policy-map POLICE-TRAFFIC
 class class-default
  police 5000000 1000000 2000000
   conform-action transmit
   exceed-action set-dscp-transmit 10
   violate-action drop

📌 Explanation:

• Traffic ≤ 5 Mbps → Allowed ✅
• Traffic between 5 Mbps – 7 Mbps → Marked as DSCP 10 ✏️
• Traffic > 7 Mbps → Dropped 🚫

---

4. Verifying Policing

✅ Check if policing is applied:

show policy-map interface GigabitEthernet1/0/1

✅ Check interface traffic rate:

show interfaces GigabitEthernet1/0/1 | include rate

---

5. Summary

Scenario	Configuration
Police all traffic to 5 Mbps	police 5000000 conform-action transmit exceed-action drop
Limit Video (AF41) to 2 Mbps	Class-based policing with exceed-action drop
Mark excess traffic instead of dropping	exceed-action set-dscp-transmit 10
Three-color policing (Transmit, Mark, Drop)	conform-action transmit, exceed-action set-dscp-transmit 10, violate-action drop
Verify policing	show policy-map interface

🚀 Traffic policing enforces bandwidth limits and protects network resources!

1. Introduction

Traffic shaping is a QoS mechanism that controls the rate of outbound traffic to prevent network congestion and packet loss. It smooths traffic bursts by buffering and delaying packets to maintain a steady transmission rate.

📌 Key Benefits of Traffic Shaping: ✅ Prevents network congestion
✅ Ensures consistent bandwidth allocation
✅ Helps with Service Level Agreements (SLAs)
✅ Reduces packet drops in bursty traffic

---

2. How Does Traffic Shaping Work?

Traffic shaping buffers excess packets in a queue instead of dropping them, then sends them at a regulated rate.

Shaping vs. Policing

Feature	Shaping	Policing
Purpose	Slows down traffic	Drops or marks excess traffic
Effect on Packets	Buffers packets in a queue	Discards or reclassifies packets
Use Case	WAN links, avoiding congestion	Enforcing strict bandwidth limits
Typical Deployment	Outbound (egress)	Inbound (ingress) & outbound

---

3. Configuring Traffic Shaping on Cisco Routers

Cisco uses Class-Based Traffic Shaping (CBTS) to configure shaping per class of traffic.

A. Shaping All Traffic on an Interface

To shape all traffic on an interface to 5 Mbps:

interface Serial0/0/0
traffic-shape rate 5000000

---

B. Class-Based Traffic Shaping

1️⃣ Define a Class to Match Traffic:

class-map MATCH-VOICE
 match ip dscp ef

2️⃣ Create a Policy Map to Shape Traffic:

policy-map SHAPE-POLICY
 class MATCH-VOICE
  shape average 1000000

3️⃣ Apply to an Interface:

interface GigabitEthernet1/0/1
 service-policy output SHAPE-POLICY

🔹 Explanation:

• Traffic matching DSCP EF (VoIP) is shaped to 1 Mbps.
• Other traffic is not affected unless added to the policy.

---

C. Configuring Hierarchical Shaping

For multiple traffic types, use Hierarchical QoS:

policy-map CHILD-POLICY
 class MATCH-VOICE
  priority 500
 class MATCH-VIDEO
  shape average 2000000

policy-map PARENT-POLICY
 class class-default
  shape average 5000000
  service-policy CHILD-POLICY

📌 This means:

• The parent policy shapes all traffic to 5 Mbps.
• The child policy shapes VoIP to 500 Kbps and Video to 2 Mbps.

---

4. Verifying Traffic Shaping

✅ To check if shaping is working:

show policy-map interface GigabitEthernet1/0/1

✅ To monitor traffic rates:

show interfaces GigabitEthernet1/0/1 | include rate

---

5. Summary

Scenario	Configuration
Shape all traffic to 5 Mbps	traffic-shape rate 5000000
Shape VoIP to 1 Mbps	shape average 1000000
Hierarchical shaping for multiple classes	Parent + Child Policy Maps
Verify shaping	show policy-map interface

🚀 Traffic shaping ensures a smooth and stable network experience by preventing congestion!

1. Introduction

The QoS Trust Boundary defines where and how QoS markings (such as DSCP or CoS) are trusted, modified, or discarded in a network. It ensures that only trusted devices (like IP phones or network switches) can set QoS values, while preventing unauthorized or misconfigured endpoints from affecting network performance.

---

2. Why is the Trust Boundary Important?

If the network blindly trusts all QoS markings:
🚨 Security Risk – End users could set high-priority DSCP values to get more bandwidth.
🚨 Misconfiguration – Incorrect markings from endpoints can cause congestion.
🚨 Unfair Bandwidth Usage – A normal PC could mark its traffic as VoIP, starving real-time applications.

To prevent these issues, network devices must determine where to trust or overwrite QoS markings.

---

3. Where Should the Trust Boundary Be Set?

Trust Scenarios

Trust Model	Description	Example Devices
Trust at the Edge	QoS markings from endpoints are accepted	Cisco IP Phones
Trust at Access Switch	Switch verifies and applies QoS policies	Cisco Catalyst Switch
Trust at Distribution/Core	Only backbone switches/routers enforce QoS	Core Routers

Best Practices

✅ Endpoints (PCs, users) – 🚫 DO NOT TRUST (Override QoS values).
✅ IP Phones, APs – ✅ Trust QoS Markings (Mark and prioritize voice traffic).
✅ Access Switches – ⚠️ Conditional Trust (Verify markings, modify if needed).
✅ Core/Distribution Layer – ✅ Strictly Enforce QoS Policies.

---

4. Configuring the QoS Trust Boundary on Cisco Switches

Cisco switches allow administrators to set trust levels on interfaces:

A. Trust DSCP from an IP Phone, Not from a PC

Most Cisco IP Phones mark their own traffic correctly, but connected PCs should not be trusted.
To trust only the phone’s QoS markings:

interface GigabitEthernet1/0/1
 switchport mode access
 switchport voice vlan 10
 mls qos trust dscp

• mls qos trust dscp → Trust DSCP values from IP phones.
• switchport voice vlan 10 → Ensures that phone traffic is correctly prioritized.

---

B. Remove Trust from an Untrusted Device (PC)

To override markings from PCs:

interface GigabitEthernet1/0/2
mls qos trust cos
mls qos trust device cisco-phone

• mls qos trust cos → Trust CoS (only if it’s coming from a phone).
• mls qos trust device cisco-phone → Trust markings only from a Cisco IP phone, not a PC.

---

C. Rewriting QoS Markings (Reclassify Traffic)

If untrusted devices send incorrect markings, we can override them:

policy-map RECLASSIFY
 class class-default
  set dscp default

interface GigabitEthernet1/0/3
 service-policy input RECLASSIFY

• This resets all traffic to DSCP 0 (Best Effort) unless explicitly classified.

---

5. Verifying Trust Settings

To check trust settings on an interface:

show mls qos interface GigabitEthernet1/0/1

To see traffic classification:

show policy-map interface GigabitEthernet1/0/1

---

6. Summary

Scenario	Configuration
Trust DSCP from IP Phone	mls qos trust dscp
Trust CoS from Cisco Phones Only	mls qos trust cos; mls qos trust device cisco-phone
Remove Trust from a PC	Apply policy to reset DSCP to default
Check Trust Settings	show mls qos interface

✅ Proper trust boundary configuration prevents abuse and ensures fair traffic prioritization! 🚀

1. Introduction

Quality of Service (QoS) ensures that critical traffic like VoIP, video streaming, and business applications get higher priority over less important traffic like bulk data transfers or web browsing.

Key Concepts

• Classification: Identifying and categorizing network traffic.
• Marking: Assigning a priority value (like DSCP or CoS) to traffic.

---

2. QoS Classification

What is Classification?

Classification identifies and groups network traffic based on parameters such as:
✅ Source/Destination IP or Port
✅ Application type (VoIP, Video, Web, FTP)
✅ Protocol (TCP, UDP, ICMP)
✅ Interface (LAN, WAN, VPN)

Traffic Classification Methods

Method	Example
Access Control Lists (ACLs)	Match traffic based on IP, protocol, or port
Class Maps	Define traffic classes in QoS policies
Network-Based Application Recognition (NBAR)	Identifies applications dynamically (Skype, Zoom, etc.)
802.1p (CoS)	Layer 2 VLAN tagging
Differentiated Services Code Point (DSCP)	Layer 3 IP header marking

📌 Example of Traffic Classification Using ACLs:

ip access-list extended VOICE-TRAFFIC
 permit udp any any range 16384 32767

📌 Using Class Maps to Classify Traffic:

class-map MATCH-VOICE
 match access-group name VOICE-TRAFFIC

---

3. QoS Marking

What is Marking?

Marking assigns priority values to packets so that network devices (routers, switches) can prioritize traffic accordingly.

Common Marking Methods

Marking Type	Layer	Bits	Value Range
802.1p (CoS)	Layer 2 (VLAN)	3 bits	0-7
DSCP (Differentiated Services Code Point)	Layer 3 (IP)	6 bits	0-63
IP Precedence	Layer 3 (IP)	3 bits	0-7

📌 Marking Traffic with DSCP (Layer 3):

class-map MATCH-VOICE
 match ip dscp 46  # Expedited Forwarding (EF) for VoIP

📌 Marking Traffic with CoS (Layer 2 VLAN):

class-map MATCH-VIDEO
 match cos 5  # High priority video

---

4. DSCP Marking Values

DSCP values define traffic priority levels:

DSCP Class	DSCP Value	Traffic Type
EF (Expedited Forwarding)	46	VoIP (Low Latency)
AF41, AF42, AF43	34, 36, 38	Video Streaming
AF31, AF32, AF33	26, 28, 30	Mission-Critical Apps
AF21, AF22, AF23	18, 20, 22	Bulk Data
CS0 (Default)	0	Best Effort

📌 Example Configuration – Mark VoIP Traffic as EF (DSCP 46):

policy-map MARK-VOICE
 class MATCH-VOICE
  set dscp ef

📌 Trust DSCP on a Switch Port:

interface GigabitEthernet1/0/1
mls qos trust dscp

---

5. Applying Classification and Marking on an Interface

1️⃣ Create an ACL to Match VoIP Traffic

ip access-list extended VOICE-TRAFFIC
 permit udp any any range 16384 32767

2️⃣ Create a Class Map

class-map MATCH-VOICE
 match access-group name VOICE-TRAFFIC

3️⃣ Create a Policy Map to Mark Traffic

policy-map MARK-VOICE
 class MATCH-VOICE
  set dscp ef

4️⃣ Apply QoS Policy to an Interface

interface GigabitEthernet1/0/1
service-policy input MARK-VOICE

---

6. Summary

Step	Command
Classify Traffic	class-map MATCH-VOICE
Mark DSCP Values	set dscp ef
Mark CoS Values	set cos 5
Apply to Interface	service-policy input MARK-VOICE

Conclusion:

• Classification helps identify network traffic.
• Marking assigns priority values (DSCP, CoS).
• Proper QoS policies ensure that voice, video, and critical data traffic get prioritized over non-essential traffic.

1. Introduction

In Quality of Service (QoS), traffic is classified and prioritized to ensure better performance for critical applications. Two common packet marking mechanisms used in QoS are:

1. IP Precedence (Legacy – 3 Bits)
2. Differentiated Services Code Point (DSCP – 6 Bits)

These values are set in the Type of Service (ToS) byte in the IP header to indicate priority levels.

---

2. IP Precedence (Legacy)

• 3-bit field (values: 0-7)
• Used in older networks (before DSCP)
• Higher values = Higher priority

IP Precedence	Binary	Priority Level	Traffic Type
7	111	Highest	Network Control
6	110	High	Internetwork Control
5	101	Critical	VoIP, Video
4	100	High	Streaming Media
3	011	Medium	Transactional Traffic
2	010	Normal	Bulk Data
1	001	Low	Scavenger Traffic
0	000	Default	Best Effort

📌 Example Configuration (Mark Traffic with IP Precedence 5 – VoIP):

class-map VOICE
 match ip precedence 5

policy-map QOS-POLICY
 class VOICE
  priority 1000

---

3. Differentiated Services Code Point (DSCP)

• 6-bit field (values: 0-63)
• More granular control than IP Precedence
• Backward-compatible with IP Precedence

DSCP Classes

DSCP values are divided into different categories:

1️⃣ Expedited Forwarding (EF – High Priority)

DSCP	Binary	Description
EF (46)	101110	VoIP, real-time apps

🟢 EF (Expedited Forwarding) is used for low-latency traffic like VoIP.

---

2️⃣ Assured Forwarding (AF – Prioritized Traffic)

Assured Forwarding (AF) provides four service classes (AF1-4) with three levels of drop probability (Low, Medium, High).

AF Class	DSCP Value	Binary	Drop Probability
AF11	10	001010	Low
AF12	12	001100	Medium
AF13	14	001110	High
AF21	18	010010	Low
AF22	20	010100	Medium
AF23	22	010110	High
AF31	26	011010	Low
AF32	28	011100	Medium
AF33	30	011110	High
AF41	34	100010	Low
AF42	36	100100	Medium
AF43	38	100110	High

🔵 Example Use Case:

• AF41 is used for video streaming.
• AF31 is used for mission-critical applications.

---

3️⃣ Default & Best Effort (Low Priority)

DSCP	Binary	Description
CS0 (0)	000000	Best Effort (Default)
CS1 (8)	001000	Background Traffic

🚫 CS0 (Best Effort) is used for general internet traffic without priority.

---

4. Mapping Between IP Precedence and DSCP

IP Precedence	Equivalent DSCP
0	CS0 (0)
1	CS1 (8)
2	CS2 (16)
3	CS3 (24)
4	CS4 (32)
5	CS5 (40)
6	CS6 (48)
7	CS7 (56)

---

5. Configuring DSCP on Cisco Devices

Marking Packets with DSCP

class-map MATCH-VOICE
 match ip dscp 46

policy-map QOS-POLICY
 class MATCH-VOICE
  priority 1000

Trust DSCP on an Interface

conf t
interface GigabitEthernet1/0/1
mls qos trust dscp
exit

---

6. Summary

• IP Precedence (Legacy): 3-bit field (0-7), basic priority levels.
• DSCP (Modern QoS): 6-bit field (0-63), more flexibility.
• EF (46) for VoIP, AF41 for Video, CS0 for Best Effort.

What is QoS?

Quality of Service (QoS) is a set of technologies and techniques used in networking to manage and prioritize traffic, ensuring efficient data transmission, reduced latency, and improved performance for critical applications.

Why is QoS Important?

Without QoS, all network traffic is treated equally, which can lead to:
✅ Poor VoIP call quality (jitter, latency, packet loss)
✅ Slow video streaming (buffering)
✅ Delayed critical applications (business or cloud apps)

QoS ensures that high-priority traffic (like voice and video) gets preferential treatment over less critical traffic (like file downloads or emails).

---

Key QoS Concepts

1. Bandwidth, Delay, Jitter, and Packet Loss

Term	Definition
Bandwidth	Maximum data transfer rate (measured in Mbps or Gbps).
Delay (Latency)	Time taken for packets to travel from source to destination.
Jitter	Variation in packet delay (problematic for VoIP and video).
Packet Loss	Percentage of lost packets, impacting data integrity.

QoS helps reduce delay, jitter, and packet loss to improve network performance.

---

2. Traffic Classification & Marking

QoS classifies packets into different categories and assigns them priorities using Differentiated Services Code Point (DSCP) or Class of Service (CoS).

Example of DSCP values:

Traffic Type	DSCP Value
Voice	EF (Expedited Forwarding – 46)
Video	AF41 (Assured Forwarding – 34)
Best Effort (Default)	0
Background Traffic	CS1 (Class Selector – 1)

To mark packets:

class-map VOICE
 match ip dscp 46

---

3. QoS Mechanisms

QoS is implemented using multiple techniques:

A. Traffic Classification & Marking

• Identifies and labels packets based on type (voice, video, data).
• Uses DSCP (Layer 3) or CoS (Layer 2).

B. Queuing and Scheduling

• Priority Queuing (PQ): Highest-priority packets are sent first.
• Weighted Fair Queuing (WFQ): Fair distribution among multiple traffic types.
• Low Latency Queuing (LLQ): Guarantees bandwidth for real-time applications like VoIP.

C. Congestion Management

• Random Early Detection (RED): Prevents congestion by dropping packets early.
• Weighted Random Early Detection (WRED): Prioritizes higher-priority packets.

D. Policing and Shaping

• Policing: Drops excess traffic above a configured rate.
• Shaping: Buffers excess traffic instead of dropping it.

Example of policing:

policy-map POLICE-TRAFFIC
 class VIDEO
  police 1000000 conform-action transmit exceed-action drop

---

QoS Configuration on Cisco Devices

1. Enable QoS on an Interface

conf t
interface GigabitEthernet1/0/1
mls qos trust dscp
exit

2. Create a Class Map (Traffic Classification)

class-map MATCH-VOICE
 match ip dscp 46

3. Create a Policy Map (Traffic Treatment)

policy-map QoS-POLICY
 class MATCH-VOICE
  priority 1000
 class class-default
  fair-queue

4. Apply QoS Policy to an Interface

conf t
interface GigabitEthernet1/0/1
service-policy output QoS-POLICY
exit

---

Conclusion

QoS is essential for ensuring a smooth network experience, especially for real-time applications like voice, video, and business-critical services. By using traffic classification, marking, queuing, and congestion management, QoS optimizes performance and prevents network slowdowns.