Skip to content
My Docs

Quality Of Service Qos

Category: Advanced Networking Concepts
Type: Network Concepts
Generated on: 2025-07-10 09:08:41
For: Network Engineering, Administration & Technical Interviews

1. Quick Overview

Section titled “1. Quick Overview”

What is QoS? Quality of Service (QoS) refers to the ability of a network to provide differentiated service to different types of network traffic. It ensures that critical applications receive the necessary bandwidth, latency, and jitter characteristics while less critical traffic is handled with lower priority.

Why is it important?

  • Improved User Experience: Ensures smooth video streaming, VoIP calls, and online gaming.
  • Business Critical Applications: Prioritizes essential applications like ERP systems or database access.
  • Efficient Resource Utilization: Optimizes network bandwidth allocation.
  • Network Stability: Prevents congestion and network degradation.
  • Meeting SLAs: Provides guarantees for service levels.

2. Key Concepts

Section titled “2. Key Concepts”
  • Bandwidth: The amount of data that can be transmitted over a network connection in a given period (e.g., Mbps). QoS can reserve bandwidth for certain traffic types.
  • Latency: The delay experienced by a packet as it travels from source to destination (measured in milliseconds). QoS aims to minimize latency for real-time applications.
  • Jitter: The variation in latency between packets. High jitter can disrupt real-time applications. QoS tries to reduce jitter.
  • Packet Loss: The percentage of packets that are lost during transmission. QoS can help minimize packet loss for critical traffic.
  • Priority: The relative importance assigned to different types of traffic. Higher priority traffic is handled before lower priority traffic.
  • Classification: Identifying different types of traffic based on various criteria (e.g., source/destination IP address, port number, application).
  • Marking: Assigning a priority tag to packets (e.g., DSCP or CoS).
  • Queuing: Placing packets into different queues based on their priority.
  • Congestion Management: Techniques to prevent network congestion, such as weighted fair queuing (WFQ) and priority queuing.
  • Congestion Avoidance: Techniques to reduce the likelihood of congestion, such as Random Early Detection (RED).
  • DiffServ (Differentiated Services): A QoS architecture that classifies and marks traffic based on DSCP values.
  • IntServ (Integrated Services): A QoS architecture that uses resource reservation (e.g., RSVP). Less scalable than DiffServ.
  • RSVP (Resource Reservation Protocol): A signaling protocol used by IntServ to reserve network resources.
  • CoS (Class of Service): Layer 2 QoS marking typically used in Ethernet networks (802.1p).
  • DSCP (Differentiated Services Code Point): Layer 3 QoS marking used in IP networks. The first 6 bits of the DiffServ field in the IP header.
  • EF (Expedited Forwarding): A DSCP value (46 or 101110) used for low-latency, low-jitter traffic (e.g., VoIP).
  • AF (Assured Forwarding): A DSCP value that provides different levels of assurance for traffic delivery. AFxy, where x is the class (1-4) and y is the drop precedence (1-3).
  • BE (Best Effort): The default traffic class with no QoS guarantees.

3. How It Works

Section titled “3. How It Works”

QoS typically involves these steps:

  1. Classification: The network device identifies the type of traffic based on criteria like source/destination IP address, port number, or application.

  2. Marking: The network device marks the packet with a QoS tag (DSCP or CoS) to indicate its priority.

  3. Queuing: The network device places packets into different queues based on their priority. Higher priority queues are serviced before lower priority queues.

  4. Congestion Management: When congestion occurs, the network device uses congestion management techniques (e.g., WFQ, priority queuing) to prioritize traffic.

  5. Congestion Avoidance: The network device uses congestion avoidance techniques (e.g., RED) to reduce the likelihood of congestion.

ASCII Diagram:

[Traffic Source] --> [Classifier/Marker] --> [Queue 1 (High Priority)] --\
                                                                        |
                                                            [Network] ---+--> [Destination]
                                                                        |
                                            [Queue 2 (Low Priority)] --/

Example Scenario:

Imagine a network with VoIP, video streaming, and web browsing traffic.

  1. Classification: The router identifies VoIP traffic based on its port number (e.g., 5060 for SIP). Video streaming traffic is identified by its destination server address. All other traffic is classified as Best Effort.

  2. Marking: The router marks VoIP packets with DSCP EF (Expedited Forwarding) and video streaming packets with DSCP AF41 (Assured Forwarding, Class 4, Low Drop Precedence). Best Effort traffic remains unmarked.

  3. Queuing: The router has three queues: EF, AF41, and Best Effort. VoIP packets go into the EF queue, video streaming packets go into the AF41 queue, and web browsing packets go into the Best Effort queue.

  4. Congestion Management: When congestion occurs, the router services the EF queue first, then the AF41 queue, and finally the Best Effort queue. This ensures that VoIP traffic has the lowest latency and jitter.

4. Protocol Details

Section titled “4. Protocol Details”
  • IP Header (DSCP):
+-------------------------------+
| Version | IHL | DSCP | ECN   |  <- DiffServ field (DSCP is the first 6 bits)
+-------------------------------+
  • Ethernet Header (CoS - 802.1p): CoS is part of the VLAN tag.
+---------------------------------+
| Destination MAC | Source MAC    |
+---------------------------------+
| 802.1Q VLAN Tag (includes CoS) | <- CoS is the PCP (Priority Code Point) - 3 bits
+---------------------------------+
| EtherType                      |
+---------------------------------+
  • No specific message flows for DiffServ, as it’s a hop-by-hop behavior. IntServ, on the other hand, uses RSVP for resource reservation. RSVP signaling involves PATH and RESV messages.

5. Real-World Examples

Section titled “5. Real-World Examples”
  • VoIP: Prioritizing VoIP traffic to ensure clear and uninterrupted calls. Using DSCP EF marking.
  • Video Conferencing: Prioritizing video conferencing traffic to prevent lagging and buffering. Using DSCP AF41.
  • Online Gaming: Prioritizing gaming traffic to minimize latency and jitter. Using DSCP AF31.
  • Financial Institutions: Prioritizing transaction processing traffic to ensure timely and accurate financial transactions.
  • Healthcare: Prioritizing medical imaging and patient monitoring traffic to ensure timely and accurate diagnosis and treatment.
  • Industrial Control Systems (ICS): Prioritizing control system traffic to ensure reliable and safe operation of industrial processes.

6. Common Issues

Section titled “6. Common Issues”
  • Misconfiguration: Incorrect DSCP markings or queue configurations can lead to unexpected results.
  • Insufficient Bandwidth: QoS cannot magically create bandwidth. If the network is already congested, QoS may not be effective.
  • End-to-End QoS: QoS needs to be implemented across the entire network path, from source to destination. A single misconfigured device can break QoS.
  • Overhead: QoS adds overhead to the network due to classification, marking, and queuing.
  • Security: QoS can be used to prioritize malicious traffic if not configured correctly. Trust boundaries are important. Don’t blindly trust DSCP markings from untrusted sources.
  • Troubleshooting: Analyzing packet captures and network device configurations is crucial for troubleshooting QoS issues.

Troubleshooting Tips:

  • Packet Captures: Use tools like Wireshark to examine packet headers and verify DSCP markings.
  • Router/Switch Logs: Check router and switch logs for QoS-related errors or warnings.
  • Performance Monitoring: Monitor network performance metrics (latency, jitter, packet loss) to identify QoS issues.
  • Configuration Verification: Double-check QoS configurations on all network devices.
  • Baseline Measurements: Establish a baseline of network performance before implementing QoS to compare against after QoS is enabled.

7. Configuration Examples

Section titled “7. Configuration Examples”

Cisco IOS Router:

! Class-map to match VoIP traffic
class-map match-any VOIP
 match protocol sip


! Policy-map to define QoS actions
policy-map QOS_POLICY
 class VOIP
  priority percent 30  ! Reserve 30% of bandwidth for VoIP
 class class-default
  fair-queue


! Apply the policy-map to an interface
interface GigabitEthernet0/0
 service-policy output QOS_POLICY

Linux (tc - Traffic Control):

Terminal window
# Create a root qdisc on interface eth0
tc qdisc add dev eth0 root handle 1: htb default 10


# Create a class for VoIP traffic (priority queue)
tc class add dev eth0 parent 1: classid 1:1 htb rate 100mbit ceil 100mbit prio 0


# Create a class for default traffic
tc class add dev eth0 parent 1: classid 1:10 htb rate 10mbit ceil 100mbit prio 1


# Filter to classify VoIP traffic (e.g., port 5060) to class 1:1
tc filter add dev eth0 parent 1: protocol ip prio 1 u32 match ip sport 5060 0xffff flowid 1:1

Important Considerations:

  • Bandwidth Allocation: Carefully plan bandwidth allocation to avoid starving other traffic classes.
  • Buffer Sizes: Adjust buffer sizes on network devices to accommodate bursty traffic.
  • Trust Boundaries: Configure QoS at the edge of the network to prevent unauthorized users from prioritizing their traffic.

8. Interview Questions

Section titled “8. Interview Questions”
  • What is QoS and why is it important? (See Section 1)
  • Explain the difference between DiffServ and IntServ. (DiffServ is more scalable and uses hop-by-hop behavior. IntServ uses RSVP for resource reservation.)
  • What are DSCP and CoS? How are they used? (DSCP is Layer 3 marking, CoS is Layer 2 marking. They are used to prioritize traffic.)
  • What is the purpose of queuing in QoS? (To place packets into different queues based on their priority.)
  • How would you prioritize VoIP traffic on a network? (Classify VoIP traffic, mark it with DSCP EF, and configure a priority queue for EF traffic.)
  • What are some common issues with QoS implementation? (Misconfiguration, insufficient bandwidth, lack of end-to-end QoS, security vulnerabilities.)
  • What is the role of congestion management and congestion avoidance in QoS? (Congestion management prioritizes traffic during congestion. Congestion avoidance reduces the likelihood of congestion.)
  • How do you troubleshoot QoS issues? (Packet captures, router/switch logs, performance monitoring, configuration verification.)
  • Explain the AFxy markings in DiffServ. (AFxy indicates Assured Forwarding, where x is the class (1-4) and y is the drop precedence (1-3).)
  • What is the significance of the EF DSCP value? (EF (Expedited Forwarding) is used for low-latency, low-jitter traffic, such as VoIP.)

Example Detailed Answer:

Question: How would you prioritize VoIP traffic on a network?

Answer: To prioritize VoIP traffic, I would follow these steps:

  1. Classification: I would identify VoIP traffic based on its characteristics. Commonly, this involves matching the SIP protocol using Layer 7 inspection or matching traffic on UDP ports 5060 and 5061 (SIP signaling) and the RTP ports (usually an even port range above 10000) for the actual voice data. I would use a class-map on a Cisco router or traffic filters (tc) on a Linux-based router.

  2. Marking: Once the VoIP traffic is classified, I would mark the packets with the DSCP EF (Expedited Forwarding) value (46 or 101110 in binary). This tells network devices along the path that this traffic requires the highest priority.

  3. Queuing: I would configure a priority queue for EF-marked traffic on the router interfaces. This ensures that packets in the EF queue are serviced before packets in other queues. On Cisco IOS, this is typically done using the priority command within a policy-map.

  4. Bandwidth Allocation: I would allocate sufficient bandwidth to the priority queue to accommodate the expected VoIP traffic volume. This prevents VoIP calls from being dropped or experiencing poor quality due to insufficient bandwidth.

  5. End-to-End Considerations: I would ensure that QoS is implemented across the entire network path, from the VoIP phones to the VoIP server. This includes configuring QoS on switches, routers, and firewalls.

  6. Monitoring and Tuning: I would continuously monitor network performance metrics, such as latency, jitter, and packet loss, to ensure that VoIP calls are experiencing good quality. I would adjust the QoS configuration as needed to optimize performance.

  7. Security Considerations: I would implement security measures to prevent unauthorized users from marking their traffic with DSCP EF. This could involve using access control lists (ACLs) or other security mechanisms.

This approach ensures that VoIP traffic receives the necessary priority to provide clear and uninterrupted calls, even during periods of network congestion.

Section titled “9. Related Concepts”
  • Traffic Shaping: Controlling the rate of traffic sent into a network to prevent congestion.
  • Policing: Limiting the amount of traffic that can be sent into a network.
  • Network Congestion Control: Mechanisms to manage and prevent network congestion.
  • MPLS (Multiprotocol Label Switching): A technology that can be used to implement QoS in carrier networks.
  • SDN (Software-Defined Networking): A networking architecture that allows for centralized control of network resources, including QoS.
  • Network Function Virtualization (NFV): Virtualizing network functions (e.g., firewalls, load balancers) to improve scalability and flexibility, including QoS capabilities.

Further Reading: