Lecture 13. Quality of Service II CM0256

Similar documents
Quality of Service II

Improving QOS in IP Networks. Principles for QOS Guarantees

Mohammad Hossein Manshaei 1393

Real-Time Protocol (RTP)

Quality of Service (QoS)

Internet Services & Protocols. Quality of Service Architecture

Real-Time Applications. Delay-adaptive: applications that can adjust their playback point (delay or advance over time).

CSE 123b Communications Software

Basics (cont.) Characteristics of data communication technologies OSI-Model

Master Course Computer Networks IN2097

Quality of Service in the Internet. QoS Parameters. Keeping the QoS. Leaky Bucket Algorithm

Multimedia Networking. Network Support for Multimedia Applications

Principles. IP QoS DiffServ. Agenda. Principles. L74 - IP QoS Differentiated Services Model. L74 - IP QoS Differentiated Services Model

Quality of Service Monitoring and Delivery Part 01. ICT Technical Update Module

Master Course Computer Networks IN2097

Quality of Service in the Internet

Presentation Outline. Evolution of QoS Architectures. Quality of Service Monitoring and Delivery Part 01. ICT Technical Update Module

Quality of Service in the Internet

Topic 4b: QoS Principles. Chapter 9 Multimedia Networking. Computer Networking: A Top Down Approach

Lecture 14: Performance Architecture

Telematics 2. Chapter 3 Quality of Service in the Internet. (Acknowledgement: These slides have been compiled from Kurose & Ross, and other sources)

CSCD 433/533 Advanced Networks Spring Lecture 22 Quality of Service

QoS Guarantees. Motivation. . link-level level scheduling. Certain applications require minimum level of network performance: Ch 6 in Ross/Kurose

A Preferred Service Architecture for Payload Data Flows. Ray Gilstrap, Thom Stone, Ken Freeman

Page 1. Quality of Service. CS 268: Lecture 13. QoS: DiffServ and IntServ. Three Relevant Factors. Providing Better Service.

Advanced Computer Networks

Lecture Outline. Bag of Tricks

Quality of Service (QoS)

Quality of Service (QoS) Computer network and QoS ATM. QoS parameters. QoS ATM QoS implementations Integrated Services Differentiated Services

Network Support for Multimedia

Institute of Computer Technology - Vienna University of Technology. L73 - IP QoS Integrated Services Model. Integrated Services Model

Design Intentions. IP QoS IntServ. Agenda. Design Intentions. L73 - IP QoS Integrated Services Model. L73 - IP QoS Integrated Services Model

Multicast and Quality of Service. Internet Technologies and Applications

EPL606. Quality of Service and Traffic Classification

Ahmed Benallegue RMDCN workshop on the migration to IP/VPN 1/54

Telematics 2 & Performance Evaluation

CS High Speed Networks. Dr.G.A.Sathish Kumar Professor EC

RSVP and the Integrated Services Architecture for the Internet

ETSF10 Internet Protocols Transport Layer Protocols

Trafffic Engineering 2015/16 1

QoS for Real Time Applications over Next Generation Data Networks

Implementing QoS in IP networks

Supporting Differentiated Services in MPLS Networks

Part1: Lecture 4 QoS

Real-Time Control Protocol (RTCP)

RSVP 1. Resource Control and Reservation

Last time! Overview! 14/04/15. Part1: Lecture 4! QoS! Router architectures! How to improve TCP? SYN attacks SCTP. SIP and H.

Advanced Lab in Computer Communications Meeting 6 QoS. Instructor: Tom Mahler

Resource Control and Reservation

Congestion Control and Resource Allocation

Telecommunication Services Engineering Lab. Roch H. Glitho

Multimedia Networking

of-service Support on the Internet

Protocols. End-to-end connectivity (host-to-host) Process-to-Process connectivity Reliable communication

Differentiated Services

2. Integrated Services

Common network/protocol functions

Lesson 14: QoS in IP Networks: IntServ and DiffServ

RSVP Petri Jäppilä Nokia Telecommunications P.O Box Nokia Group, Finland

Multi-Protocol Label Switching

QUALITY OF SERVICE ARCHITECTURES APPLICABILITY IN AN INTRANET NETWORK

CS 268: Integrated Services

Protocols for Multimedia on the Internet

Signalling Overview. IP Precedence

Author : S.chandrashekhar Designation: Project Leader Company : Sasken Communication Technologies

ip rsvp reservation-host

Modular Quality of Service Overview on Cisco IOS XR Software

Quality of Service Basics

QoS in IPv6. Madrid Global IPv6 Summit 2002 March Alberto López Toledo.

Week 7: Traffic Models and QoS

EE 122: Differentiated Services

Internet Engineering Task Force (IETF) December 2014

Quality of Service In Data Networks

IntServ and RSVP. Overview. IntServ Fundamentals. Tarik Cicic University of Oslo December 2001

Internet Quality of Service: an Overview

DiffServ Architecture: Impact of scheduling on QoS

QoS Requirements and Implementation for IMS Network

DiffServ over MPLS: Tuning QOS parameters for Converged Traffic using Linux Traffic Control

Towards Service Differentiation on the Internet

Quality of Service. Qos Mechanisms. EECS 122: Lecture 15

MultiProtocol Label Switching - MPLS ( RFC 3031 )

Quality of Service for Multimedia over Next Generation Data Networks

Today. March 7, 2006 EECS122 Lecture 15 (AKP) 4. D(t) Scheduling Discipline. March 7, 2006 EECS122 Lecture 15 (AKP) 5

Marking Traffic CHAPTER

ITBF WAN Quality of Service (QoS)

Quality of Service (QoS)

Improve the QoS by Applying Differentiated Service over MPLS Network

A MPLS Simulation for Use in Design Networking for Multi Site Businesses

Quality of Service In Data Networks: Problems, Solutions, and Issues

Multiplexing. Common network/protocol functions. Multiplexing: Sharing resource(s) among users of the resource.

Configuring Quality of Service for MPLS Traffic

QoS Technology White Paper

DiffServ over MPLS: Tuning QOS parameters for Converged Traffic using Linux Traffic Control

Master Course Computer Networks IN2097

Master Course Computer Networks IN2097

QoS Technology White Paper

Internet QoS : A Big Picture

Grandstream Networks, Inc. GWN7000 QoS - VoIP Traffic Management

INTEGRATED SERVICES AND DIFFERENTIATED SERVICES: A FUNCTIONAL COMPARISON

University of Cyprus Computer Science. Implementation and Evaluation of Differentiated Services on Linux

Transcription:

Lecture 13 Quality of Service II CM0256

Types of QoS Best Effort Services Integrated Services -- resource reservation network resources are assigned according to the application QoS request and subject to the bandwidth management policy Differentiated Services -- prioritization the network traffic is classified and network elements give preferential treatment to classifications identified having more demanding requirements.

Integrated Services Integrated Services Internet Architecture particular architecture for providing QoS Intserv is a framework developed within the IETF to provide individualised quality of service guarantees to individual application sessions it has two key features: 1. Reserved Resources: a router is required to know how many of its resources (buffer, link bandwidth) are already reserved for on-going sessions 2. Call Setup: a session requiring QoS guarantees must first be able to reserve sufficient resources at each network router on its source-to-destination path to ensure its end-to-end QoS requirement is met Call setup (or call admission) must involve all routers along a path from source-todestination to ensure QoS is met. This process must therefore consider resources (buffers, links, etc) at each router along the chain

Call Admission Process Traffic characterisation and specification of the desired QoS: for a router to determine whether or not its resources are sufficient to meet the QoS requirements of a session, that session must first declare its QoS requirement, as well as characterise the traffic that it will be sending into the network, and for which it requires a QoS guarantee Rspec: In Intserv architecture defines the specific QoS being requested by a connection R is for reserved Tspec: In Intserv architecture characterises the traffic the sender will be sending into the network, or the receiver will be receiving from the network T is for traffic Signalling for call setup: A session s Tspec and Rspec must be carried to the routers at which resources will be reserved for the session RSVP is the signalling protocol of choice (see later) Per-element call admission: Once a router receives the Tspec and Rspec for a session requesting a QoS guarantee, it can determine whether or not it can admit the call this decision is based on the existing commitments, traffic specification, and the requested type of service

Intserv: Service Classes The IntServ architecture adds two service classes to the existing best effort model: Guaranteed Quality of Service (RFC 2212): provides mathematical bounds on the queueing delay a packet will encounter at a router. Traffic characteristics modelled by a leaky bucket flow with parameters (r,b) Control Load Network Service (RFC 2211): such a session will receive a QoS closely approximating the QoS that same flow would receive from an unloaded network element. Hence, assume a very high success in forwarding packets, with very little packet loss (assume that this is the only traffic) The Controlled Load service targets real-time multimedia applications that have been developed for today s internet

ReSerVation Protocol (RSVP) To provide QoS guarantees, a signalling protocol is needed to enable resource reservation in the Internet RSVP is such a protocol (RFC 2205) Resources are generally link bandwidth and router buffers ( in the context of Internet ) In essence, RSVP enables a host to reserve bandwidth on behalf of an application data flow and is used by routers to forward these bandwidth requests Hence, RSVP software must be present in the receivers, senders and the routers RSVP has two principal characteristics: 1. Provides reservation for bandwidth in multicast requests 2. It is receiver-oriented the receiver of the data flow initiates and maintains the resource reservation used for that flow RSVP is a signalling protocol allows hosts to establish and tear down reservations for data flows terminology comes from circuit switching

RSVP Session In RSVP, a session can include multiple (multicast) data flow streams. Each sender is the source of one or more data flow such as a video or audio flow where each data flow has the same multicast address Routers and hosts identify the session to which a packet belongs by the packet s multicast address (a simplification of reality RSVP allows more general methods to identify a session) Within a session, the data flow to which a packet belongs also needs to be identified

RSVP Does Not Specify how the network provides the bandwidth to data flows Provide routing does not determine the links the reservations are to be made utilises an underlying routing protocol (unicast or multicast) to determine the routes for the flows RSVP merely enables reservation of link bandwidth once these reservations are in place, it is up to the routers to provide the reserved bandwidth to the data flows Achieved with scheduling mechanisms such as priority, scheduling and weighted fair queue as seen earlier When a route changes RSVP re-reserves resources RSVP is therefore one piece in the QoS puzzle

Heterogeneous Receivers Receivers can accept flows at different rates, such as 28.8 kbps, 128 kbps or 10 Mbps should a sender encode for the lowest rate or for the highest rate? Video and audio is encoded in layers the base layer could have a rate of 20 kbps, and an enhancement layer could have a rate of 100 kbps Hence, the receivers with different rates could receive different layers to construct a quality image/sound locally In this case, the sender does not need to know receiving rates of all receivers only the maximum rate of all its receivers the receivers then automatically pick up suitable layers from the multicast stream Receivers must communicate to the network the rates they can handle via a reservation message Specialised Path Messages let routers know the links on which they should forward the reservation messages from senders to receivers

Reservation Style Reservation style specifies whether merging of reservations from the same session is permissible Also specifies session senders from which a receiver desires to receive data a router, for instance, can determine the sender of a datagram from its source IP address Three different reservation styles: 1. Wildcard-filter style: When a receiver uses this style, it is telling the network that it wants to receive all flows from all upstream senders in the session, and that bandwidth reservation is to be shared between senders 2. Fixed-filter style: Receiver specifies a list of senders from which it wants to receive a data flow along with a bandwidth reservation for each of these senders. These reservations are not to be shared 3. Shared-explicitly style: Receiver specifies a list of senders from which it wants to receive a data flow along with a single bandwidth reservation request. This reservation request is to be shared between all the senders in the list (1) and (3) are appropriate for a multicast session where sources are unlikely to transmit simultaneously e.g. packetised speech (not everyone talks at the same time) (2) is appropriate for video-conferencing

Transport of Reservation Messages RSVP messages are sent hop-by-hop over IP hence RSVP is placed in the information field of the IP datagram (protocol number set to 46) RSVP messages can also get lost, as IP is unreliable a replacement refresh should be sent if this happens RSVP reservation message has IP of source and IP of destination multicast router in its IP datagram when received at router the IP fields are stripped, and handed to the RSVP module of the router This module examines the style type, its multicast address (session identifier), current state, and then acts appropriately for instance, it may merge this reservation with a reservation from another interface, and send a new reservation message to next router upstream

Insufficient Resources Since a reservation request may be a merged request, a reservation error must be reported to all concerned receivers use of ResvError messages Receivers can then reduce the amount of resources they request and try again RSVP can facilitate backtracking of the reservations when suitable resources are not available If a large request from one receiver are merged with lots of small requests from other receivers, it is possible for the large request to override the small ones Blockade state block out offending receiver always requesting large amount of resources repeatedly, and preventing other receivers from getting their small requests RSVP is quite complicated!

Intserv / RSVP problems Based on RSVP per-flow resource reservation, it is possible to provide QoS guarantees to individual flows Intserv and RSVP based reservation however have problems: Scalability: routers must maintain state information for each flow passing through it. This is a considerable overhead Flexible Service Model: Intserv framework only provides a small number of pre-specified service classes for providing relative services. This is restrictive, as more qualitative or relative definitions cannot be provided by an application to better reflect the application demands

Differentiated Services (Diffserv) Some of the problems of Intserv/RSVP are being handled by the IETF in the Diffserv model such as scalability and flexible service differentiation Aims to provide the ability to handle different ways within the Internet Diffserv does not pre-define classes but provides functional components with which such services can be built To support a large number of simultaneous flows within the router backbone. Diffserv aims to provide only a small subset of functionality within the network core with more complex control functions being supported at the edge of the network RFC 2474/2475 define the fundamental framework for Diffserv

Diffserv Traffic Classification and Conditioning Diffserv field carried within a packet (IPv6 or IPv4 header) replaces the Type-of-Service (ToS) field The 6 bit Differentiated Service Code Point (DSCP) defines how the packet behaves at every hop within the network this acts as a mark to identify a particular class for the packet This field is set before the packet enters a network (at network edge) either at a Diffserv capable host or a Diffserv capable router Packets arriving at a Diffserv capable host or router are first classified, based on their source/destination address, source/destination port, protocol ID etc - and then, based on this, are steered to the appropriate marking function The classification is based on pre-defined rules defined either by a network administrator, or a yet-undefined signalling protocol

Diffserv - II At each subsequent Diffserv capable router along the chain to the destination, packets are given service based on their marks Packet marking could be achieved on a number of complicated schemes related to number of packets sent from a source (traffic profile of source), a negotiated traffic profile (monitored via a metering function) Diffserv does not mandate any specific policy for what marking and conditioning (shaping) is actually to be achieved with the sent packets

Diffserv III Per-Hop Behaviour Per-Hop Behaviour (PHB) is the second important aspect of Diffserv architecture, after edge behaviour PHB involves a number of considerations: 1. A PHB can result in different classes of traffic receiving different performance (i.e. have a different externally observable forwarding behaviour in Diffserv terminology) 2. PHB only defines differences in performance (behaviours) between classes, not how these performance differences are to be achieved hence any implementation mechanism will suffice, provided the externally specified behaviour is achieved 3. PHB could involve different buffer/bandwidth allocation schemes 4. Differences in performance must be observable and therefore measurable

Per-Hop Behaviour Examples Provide a guarantee that a given class mark packets receive at least x% of the outgoing link bandwidth over some interval of time One class of traffic will always receive strict priority over another if both high-quality and low-quality packets are present in router queue at the same time for instance Two PHBs: 1. Expedited Forwarding (EF): departure rate of a class of traffic from a router must equal or exceed a configured rate. This essentially means that a given class of traffic must be guaranteed enough bandwidth to meet or exceed a given minimum rate 2. Assured Forwarding (AF): divides traffic into four classes, where each AF class is guaranteed to be provided with some minimum amount of bandwidth and buffering. In each of the four classes, there are three additional drop preference categories. If congestion occurs within an AF class, a router can then discard (drop) packets based on their drop preference values (RFC 2597) EF implies some level of isolation, as guarantee is made independently of traffic intensity of other classes arriving at a router AF per-hop behaviour can be used to provide different levels of service based on drop preference values Traffic demands and rate of a given service class must be taken into account, in addition to resource reservation requests for a given class when considering per-hop behaviour

Multiprotocol Label Switching (MPLS) MPLS is similar to Diffserv, as it marks traffic at ingress boundaries in a network, and un-mark at egress points. But unlike Diffserv, which uses the marking to determine the priority within the router, MPLS markings (20-bit labels) are primarily designed to determine the next router hop. MPLS resides only on routers MPLS is a protocol-independent, it can be used with network protocols other than IP (like IPX, ATM, PPP or Frame-Relay) or directly over datalink layer as well.

MPLS - I Ingress router Egress router [From Cisco white paper]

MPLS - II MPLS simplifies the routing process (decreases overhead to increase performance) The process used by the MPLS-enabled routers is called Label Switching Route (LSR). It functions as follows: At the first hop router in the MPLS network (ingress router), the router makes a forwarding decision based on the destination address or any other information in the header, as determined by local policy, then determines the appropriate label value which identifies the Forwarding Equivalence Class (FEC) attached the label to the packet and forwards it to the next hop At the next hop, the router uses the label value as an index into a table that specifies the next hop and a new label. The LSR attached the new label, then forwards the packet to the next hop. The router taken by the MPLS-labelled packet is called Label Switched Path (LSP). Advantages of using MPLS: routers have less work to do and can act more like simple switches