Network Design - Chapter 2: Business and Technical Goals and Requirements - University of Pittsburgh

Nội dung text: Network Design - Chapter 2: Business and Technical Goals and Requirements - University of Pittsburgh

1. Business and Technical Goals and Requirements David Tipper Associate Professor Department of Information Science and Telecommunications University of Pittsburgh Slides 2 Last Week • Network Design is not a precise science. – Many different types of problems • greenfield vs. incremental, • LAN vs. MAN vs WAN • wired vs. wireless •Etc. – There can be many good answers - no best solution – Design involves trade-offs among cost vs. performance • Top Down Design approach useful as a framework – Conceptual Model – Logical Model – Physical Model TELCOM 2110 2 1
2. Top Down Network Design • Top Down Network Design – Conceptual Model •Objectives – Business Goals, Technical Goals • Requirements – Business (e.g.,support XYZ application), Technical (availability, delay, bandwidth, security, etc.,) • Constraints – Business (organizational, budget, etc.,) and Technical (vendor, technology, sites to connect, etc.) – Logical Model • Technology, network graph, node location, link size, etc. (where algorithms are used to minimize cost) – Physical Model • Specific hardware/software implementations • (e.g., wiring diagram, repeater locations, etc.) TELCOM 2110 3 Conceptual Model Design Objectives • Identify Business Objectives of the network design project – Through surveys/questionnaires, meetings • What will the network be used for? Mission critical uses? • How does the customer think the new network will improve their business practices? • What is the criteria to be used to judge the network success/failure? • Example objectives – Provide new services/Modernize out-dated technologies – Reduce network costs – Make more data available to more people – Improve network security and reliability TELCOM 2110 4 2
3. Business Goals • What is the scope of the project? – Greenfield or incremental design – Network Size • (LAN, Campus net, enterprise WAN, backbone, etc.) • Sites to be connected, distances, etc. – What are current/existing networks and services – Identify applications and services to be provided from surveys and meetings – Quantify user behavior – As a guideline construct a table with the following info Type of New App? Application Application Freq of use Criticality Comments Sales Distributed No/hourly Very Tracking client/server TELCOM 2110 5 Application Modeling • Need to quantify application behavior • Can roughly classify applications into categories – Distributed Client Server – Cooperative Computing – Distributed Computing – Peer-to-Peer Model • Classifications used to define – application flows directions and characteristics • Unidirection or bidirectional • Symmetric or asymmetric • low, medium, or high bit rate – flow boundaries • e.g., LAN-WAN traffic, intra- campus flow etc. TELCOM 2110 6 3
4. Application Types • Distributed Client Server – Tend to be produce asymmetric traffic – larger in downstream direction – Hierarchical organization – Examples: • Web applications • Sales Tracking etc. TELCOM 2110 7 Application Types • Cooperative Computing – Tend to be produce asymmetric traffic – larger in downstream direction – Managed correlated distributed servers – Variation of client server – Examples: • Collaborative Document Processing • Inventory Control/management TELCOM 2110 8 4
5. Application Types • Distributed Computing – Tend to be produce symmetric traffic – Managed correlated computers – Examples: • Computer Aid Manufacturing • Computer Aided Design TELCOM 2110 9 Application Types • Peer-to-Peer Model – No obvious hierarchy or asymmetry to traffic – Examples: • ftp, telnet • Video/audio conferencing TELCOM 2110 10 5
6. Applications Map • List Applications supported at various sites and between sites • Example- company with offices in Dallas and Vienna, VA, • Factory in Denver- consider WAN applications only • Appl A: Sales/inventory control • Appl B: CAM • Appl C: CAD • Appl D: video conference • Appl E: Intranet Voice over IP TELCOM 2110 11 Applications Map • From Applications Map – get rough idea of traffic flows between network nodes • Get the beginnings of a traffic demand matrix across the network – For example – applications across the WAN table below TELCOM 2110 12 6
7. Characterizing Application/ Network Traffic • Traffic Characterization – What kind of traffic is generated? • Client-server, peer-to-peer, etc. – How often is it generated? – What is the relative impact on the network? • Peak rate, mean sustained rate, min rate • Burst size, burst duration • Degree of multi-casting • If possible characterize data network traffic on existing network – two options – Application Monitoring • What are applications and how much bandwidth needed – Network Monitoring • What is network traffic pattern and bandwidth usage TELCOM 2110 13 Characterizing Network Traffic • Application Monitoring – If incremental network design and current application – benchmark traffic • Software tools can be used to determine application performance statistics • Uses “agents” to collect data and send information to a “management” station • Agents run on the different OS where the applications are installed • Standalone software or integrated into network management software (e.g., Openview) • Normally, the profiling software transforms raw application data captured from the network into an application profile TELCOM 2110 14 7
8. Application Monitoring Software TELCOM 2110 15 Application Monitoring Software TELCOM 2110 16 8
9. Application Usage Patterns • Application Monitoring allow one to develop Profiles of each Application – Number of users, – Number of sessions per user-day – Average duration of session – Average number of simultaneous sessions – Peak data rate, Burst Duration – Mean data rate, min data rate, multi-cast, etc. • Translate Application monitor and profile data into traffic demand matrix TELCOM 2110 17 Application Assumptions • If greenfield design or unable to benchmark network – assume characteristics from similar applications or other benchmark studies • Can use data gathered from user surveys or make overly conservative assumptions : – number of application users = # simultaneous users – all applications are used all the time – each user opens just one session and the session lasts all day – Typical values of applications data given in textbook • Web page with graphics - 50 Kbytes • Spreadsheet - 100 Kbytes • Word processing document - 200 Kbytes TELCOM 2110 18 9
10. Network Monitoring • Identify Traffic Flows – Establish traffic flow boundaries • Host to server • Floor – to – floor • LAN to WAN • Management traffic • Multi-cast •Etc. – Capture the appropriate traffic for each flow – Use a network capturing and analysis tool • Sniffer, Network Management software, etc. – Identify each flow in the capture – Can separate flow by protocol type, destination, etc. TELCOM 2110 19 Network Monitor Software Network Monitor - [\Ethernet\NET1 Capture Window (Station Stats)] File Capture Tools Options Window Help % Network Utilization: Time Elapsed: 00:01:44.659 00 100 Network Statistics Frames Per Second: # Frames: 35 Graph Pane # Broadcasts: 4 Graph0 Pane 0 100 # Multicasts: 0 Bytes Per Second: # Bytes: 3450 Total 00 2180 # Frames Dropped: 0 Total Broadcasts Per Second : Network Status: Normal Statistics Pane Captured Statistics Statistics Pane # Frames: 35 Network Address 1->2 1<-2 Network Address 2 # Frames in Buffer: 35 BACKUP 9 11 WFW Client # Bytes: 3450 Session # Bytes in Buffer: 3730 SessionBACKUP 1 *BROADCAST % Buffer Utilized: 0 StatisticsStatisticsINSTRUCTOR PanePane 2 1 WFW Client # Frames Dropped: 0 INSTRUCTOR 4 4 BACKUP Per Second Statistics WFW Client 3 *BROADCAST % Network Utilization: 0 # Frames/second: 0 # Bytes /second : 0 Network Address Frames Sent Frames Rcvd Bytes Sent Bytes Rcvd Directed Frames Sent Multicasts Sent Broadcasts Sent *BROADCAST 0 4 0 423 0 0 0 BACKUP 14 15 1336 1513 13 0 1 StationStation INSTRUCTOR 6 5 432 402 6 0 0 Statistics Pane WFW Client 15 11 1682 112 12 0 3 StatisticsStatistics PanePane Network Monitor V1.1 (built on Jun 23 1995 at 17:49:57) TELCOM 2110 20 10
11. Network Monitoring • Method for Characterizing a Traffic Flow – Determine statistics for traffic flows • Individual flow, composite flow, backbone flow – Peak data rate, burst duration, mean data rate, mean response time, etc. – Create source –destination traffic matrix – May include path info in matrix Destination 1 Destination 2 Destination 3 Destination 4 Source LAN 3 Mbps 500 Kbps/Path 100 Kbps 50 Kbps Segment 1 A-B-C TELCOM 2110 21 Example • Example- company with offices in Dallas and Vienna, VA, • Factory in Denver • Appl A: Sales/inventory control •ApplB: CAM •ApplC: CAD • Appl D: video conference • Appl E: Intranet Voice over IP TELCOM 2110 22 11
12. Applications Map • From Applications Map – get rough idea of traffic flows between network nodes • Get the beginnings of a traffic demand matrix across the Wide Area Network • If use Applications Monitoring Approach –gather data on each application • A: Mean rate = .1 Mbps, Peak = .15 Mbps • C: Mean rate = .5 Mbps, Peak = .75 Mbps • D: Mean rate = 2 Mbps, Peak = 2.5 Mbps TELCOM 2110 23 Traffic Demand Matrices • From the application map and associated matrix and the application monitoring data we have the mean traffic demand matrix and peak traffic demand matrix as below • Note, if the network monitoring approach is used get traffic demand directly. Mean data Dallas Denver Vienna Peak data Dallas Denver Vienna rate demands rate demands Dallas .1Mb 2.1 Mb Dallas .15Mb 2.65 Mb Denver .3 Mb .8Mb Denver .45 Mb 1.2Mb - Vienna 2.5 Mb 1.5 Mb Vienna 3.25 Mb 2.25 Mb TELCOM 2110 24 12
13. Technical Requirements & Constraints • From surveys/questionnaires, meetings etc. application data determine technical requirements and constraints • Technical goal is to build a network that meets user’s requirements + some they may not know they need. • Technical Goals – Scalability – Availability/reliability – Network Performance • Utilization, Throughput, Delay, Delay Jitter, packet loss rate, call/connection blocking rate • Traffic Estimation crucial –Security – Manageability/Interoperability – Affordability $$ • Need to determine reasonable goal for each category and the importance of each. TELCOM 2110 25 Scalability • Scalability – how much growth a network design can support – can the design adapt to changing network load and QoS requirements? – Need to examine the network needs out a few years • Key points to understand – How many more sites will be added? – How extensive will networks be at each site? – How many more users will be added? – How many more servers, etc will be added? – How many and what applications will be added? – Technology migration path? TELCOM 2110 26 13
14. Scalability • Scalability – For logical network design – how much additional traffic can be added – without substantial additional investment – For physical design - thought of as expandability and upgrade capability – For example, • Given specific Router • Can interface bit rate be upgraded • Can number of I/O ports be increased? • Can additional software features be added (e.g, VLAN capability, IP Sec etc.) – Try to set reasonable scalability goals TELCOM 2110 27 Availability • Availability is the amount of time a network is available to users • Can be expressed as percent uptime – 165 hours in 168 hours/week = 98.21% • Redundancy is used to increase a networks availability • Availability Goals depend on application and user requirements – may vary with location – Highly available voice service through out company – traditionally telephone systems have five 9s availability –> 99.999% => – downtime = .00001 x 60 x 24 x 365 = 5.25 minutes per year! TELCOM 2110 28 14
15. Availability Goals Availability level Downtime per year 99.999% 5.25 min 99.97% 157.68 min 99.9% 8 hours 46 min 99% 87 hours 4 min •Work with customer to set realistic availability goals • Telcom equipment traditionally five 9 availability • Data equipment usually two or three 9 availability • Remember System Availability is less than component equipment availability TELCOM 2110 29 System Availability Goals • System availability A estimated from component availability Ai • If devices in series N AAseries= ∏ i i =1 N A parallel=−1(1)∏ −A i i =1 TELCOM 2110 30 15
16. Availability Goals A = .99999 x (1 - (1-.999)(1 -.999)) x (1 – (1-.99) (1-.99)) High availability requires increased cost $ Need increased component availability and redundancy via parallel routes TELCOM 2110 31 Network Performance Goals • There are several measures to look at – Utilization – Throughput – Accuracy (Packet Loss) – Efficiency • Protocol Frame Size, Protocol Overhead, and Routing Protocol Overhead – Delay and Delay Jitter – Call Blocking for circuit switched networks • Typically look at measures during the busy period of the day set threshold values TELCOM 2110 32 16
17. Network Performance • Typically have a camelback shape to network traffic (both packet and circuit switched networks) • Busy time period will vary with network application (e.g. is commuting time in urban cellular networks) TELCOM 2110 33 Network Performance - Utilization • Utilization is the percent of total available capacity (bandwidth) on a link in use (0-100%) • Bandwidth utilization is measured over a time interval to determine the amount in use (e.g. the busy hour or some fraction of it) • Link/equipment utilization identifies network bottleneck points – Data networks have utilization < 40 - 60% – Telephone network utilization 80 – 90 % • Utilization goals will effect resulting delay TELCOM 2110 34 17
18. Network Performance - Throughput • Throughput is defined as the quantity of error- free data successfully transferred between nodes per unit of time (Goodput or Layer 2/3 throughput) • Depends on network access method, the load on the network and the error rate • Throughput can be expressed – in Packets per Second (PPS) than can be sent by a device with dropping any packets for data networks or bps – Carried load in Erlangs for circuit switched networks – Example IEEE 802.11 b wireless LAN – channel rate 11 Mbps – typical throughput 7 Mbps TELCOM 2110 35 Network Performance -Accuracy • Accuracy is a measure to ensure that the data received at the destination must be the same as the data sent by the source • Data errors are caused by power surges, or spikes, poor physical connections, failing devices and electrical noise • Accuracy can be expressed in Bit Error Rate (BER) or packet error rate (PER) • Target values of BER depend on physical medium used wireless link – 1 in 104 , optical fiber 1 in 1010 TELCOM 2110 36 18
19. Network Performance -Accuracy • Packet Loss occurs when buffers overflow at routers or gateways • Packet Loss results in retransmission in applications that require reliability • In real-time applications retransmission is not an option after packet loss • Some low level of packet loss can be made up by human brain from context in audio/video • Quality drops quickly with increasing packet loss rate • For quality comparable to PSTN need very low loss rate < 0.5% • Packet Loss increase is highly nonlinear with load increase TELCOM 2110 37 Network Performance -Efficiency • How much “overhead” is needed to send traffic across the network • Overhead is due to several factors lets look at some of them: – Packetization Overhead – Network Protocol Overhead – Routing Protocol Overheads • Remember data is packaged in protocol frames that contain overhead data, some have more overhead than others – Ethernet - 38 bytes per frame – IP - 20 bytes per frame – TCP - 20 bytes per frame – ATM - 5 bytes per cell – IP RIP - every 30 seconds sends 532 byte packets • Overhead effects delay and link sizing • Example VoIP (IP/UDP/RTP) IP packet Header UDP packet Header 20 Bytes Header RTP packet 8 Bytes 12 Bytes Data payload TELCOM 2110 38 19
20. Network Performance - Delay • Interactive applications demand minimal delay when receiving a data stream • Delay must be constant for real-time applications like voice and video applications other wise you will get jitter causing disruptions in voice quality and jumpiness in video streams • Delay Jitter is the variability in the delay from a constant • Delay caused by network devices that move the data within a network (e.g., router) • For example consider Voice over IP TELCOM 2110 39 IP Telephony Delays •Consider VoIP only network (no gateways or PSTN) •Major Delays in IP Telephony Systems •Coding •Packetization/Serialization •Queueing at Routers •Propagation •Dejitter •Decoding TELCOM 2110 40 20
21. IP Telephony Delays • Coding Delay • Time to gather speech sample compute vocoder model values for transmission • Value depends on vocoder utilized (0-50ms) • Packetization and Serialization • Packetization: Time to gather data from coder for packet payload, attach headers • Remember the protocol stack for VoIP Output of Vocoder packed in Real Time Protocol (RTP) packets Which are payload for User Datagram Protocol (UDP) packets Which are payload for Internet Protocol packets (IP) TELCOM 2110 41 Packetization Delay • VoIP packet (RTP/UDP/IP) total header = 40 Bytes IP packet Header UDP packet Header 20 Bytes Header RTP packet 8 Bytes 12 Bytes Data payload • Assume –Delay: N voice samples → T ms -> payload P – Payload efficiency: P/(P+Header) % – Net data rate: (P+Header)/T = R Kbps TELCOM 2110 42 21
22. Packetization and Delay Data stream Accumulation (Compressed) Buffer delay IP packet Header UDP packet Header 20 Bytes Header RTP packet 8 Bytes 12 Bytes Data payload • For example: 10Byte payload from 4-to-1 compression rate vocoder –Delay: 10Byte → 40 samples → 40×125μs = 5ms – Packet efficiency: 10/(40+10) = 20% – Net data rate: 50B/5ms = 80 Kbps (>64 kbps DSO!) TELCOM 2110 43 Serialization and Transmission • Serialization Delay: time to transmit on access line both from caller to network also have this at the other end of network to called party – 1 byte on 64kbps line => 125 μsec – G.723a VoIP codec over modem: 64byte packet /56kbps=11ms – 1byte on OC-3 optical fiber to home line (155Mbps) => 0.05 μsec – Insignificant on high-speed links • Propagation Delay – Time to propagate packet down link - depends on distance of link and medium • Satellite Hop wireless link 250 ms • Coast to Coast in North America fiber optic propagation 24 ms • For example fiber optic cable propagates at roughly 2/3 speed of light (3 x 108 ) meter/sec - so 200km link has propagation delay of 200/(3 x 108 ) = 0.66 ms – Small enough on short fiber links to ignore TELCOM 2110 44 22
23. Network Delays • Router delay – Time for router to process/transmit packet + delay in router queues – Time to process/transmit packet depends on router switch speed and link speed – for high bandwidth links and core network routers small amount of time 10 – 20 μsecs Queueing Delay 25 –Time waiting in router buffers for 20 processing and transmission –Value highly dependent on load and 15 QoS mechanisms deployed in router 10’s msec to 10’s secs 10 •Queueing Delay nonlinear with 5 increases of network load 0 0 .6 .8 0.1 0.2 0.3 0.4 0.5 0 0.7 0 0.9 TELCOM 2110 45 Network Delays • Delay Jitter defined as the variation of the delay for two consecutive packets • Due to variation of – Routes of packets – Router delay (processing time + queueing time) TELCOM 2110 46 23
24. Network Delays • Jitter buffer – Jitter buffer to smooth out playout of packets to destination • Allows packet delivery times to vary • Allows packets to arrive out of order – Note 30 ms holds one G.723 packet, typical values 30-100 msec Receive Buffer CODEC Jitter eliminated if buffer is sufficiently large TELCOM 2110 47 Example of End-to-End Delay Budget • Often design on basis of a Target Delay Budget • Sender – Coding Delay 5 – Packetization delay 30 – Serialization delay 11 •Network If no congestion. – Routers 5 @ 7ms each 35 – Propagation 25 • Receiver – Jitter buffer 30 – Serialization, de-packet, decode 46 • Total 182 ms • Well below 400ms but above VoIP ideal of 150ms TELCOM 2110 48 24
25. Network Performance - Response Time • Response time is a network performance goal that users care about most • Users recognize the amount of time to receive a response from the network system • Users begin to get frustrated when response time is 100ms (.1 seconds) or greater TELCOM 2110 49 Call Blocking /Traffic Engineering • In circuit switched networks – the main performance metric is call blocking • Traffic Engineering or Trunking is term for determining the call blocking and capacity allocation • Required grade of service? – Usually 1% blocking probability during busy hour – Busy hour may be (1) busy hour at busiest switch (2) system busy hour (3) system average over all hours • Estimated traffic distribution? – Traffic intensity is measured in Erlangs (mathematician AK Erlang) – One Erlang = completely occupied channel, • For example, a standard DS0 64Kbps telephone channel occupied for 30 min. per hour carries 0.5 erlangs TELCOM 2110 50 25
26. Traffic Engineering TELCOM 2110 51 Traffic Engineering • Traffic intensity per user Au = average call request rate λ x average holding time H • Total traffic intensity = traffic intensity x number of users = Au x nu • Given T traffic channels - what is GoS or how many users can be supported for a specific GoS? • Basic analysis same for all circuit switched telephony (wired or wireless), – Erlang B model TELCOM 2110 52 26
27. Erlang B Model M/M/C/C queue • C identical servers process customers in parallel. • Customers arrive according to a Poisson process • Customer service times exponentially distributed • The system has a finite capacity of size C, customers arriving when all servers busy are dropped • Blocked calls cleared model (BCC) • Analyze using Markov Process of n(t) – number of customers in the system at time t μ λe = λ(1− Pb) λ λe λPb TELCOM 2110 53 Erlang B Charts TELCOM 2110 54 27
28. Traffic Engineering Erlang B Table TELCOM 2110 55 Traffic Engineering Erlang B table TELCOM 2110 56 28
29. Security • Security design is becoming one of the most important aspects of network design • Network design must ensure against loss of business data or disruption of business activity • Need to understand the risk of data loss • Security Concepts – COMMSEC: security at communications level – INFOSEC: security at information level TELCOM 2110 57 Security Threats • System Intrusion – Improper access to network and hosts resources • Denial of service – Disable network and hosts • Snooping • Spoofing • Data manipulation • Physical damage • Information Assurance – info security + info availability TELCOM 2110 58 29
30. Security Impact on Network • Security Mechanisms must be put in place to provide security – Physical Security Measures • Servers/cabling in locked rooms • Backup power and storage, etc • Impacts physical design – Electronic Security Measures • Authentication, packet filters, encryption • Firewalls – Impacts network performance => > capacity TELCOM 2110 59 Manageability • There are different ways to manage a network and the different things to manage – Performance, security, fault, configuration, accounting, etc. • Management architecture needs to be deterimined – In-band versus out-of-band monitoring/signaling – Centralized vs. distributed monitoring – Estimate additional traffic due to management flows • Also need to consider interoperability with existing infrastructure and management TELCOM 2110 60 30
31. Affordability • Affordability is sometimes called cost- effectiveness • Want to carry the maximum amount of traffic for a given financial cost • Financial costs include non-recurring equipment costs and recurring network operating costs • Campus, Metro and WAN costs are areas where a good design can save $ TELCOM 2110 61 Ranking • Useful to have users/management rank performance goals – Low delay more important than availability – Ease of management more important than security – Comparative ranking or absolute – One approach is assume 100 point to be distributed among the categories of interest and users must allocate the points among the performance categories • (scalability, availability,delay, security, etc.) • Provides Guidance to optimizing network design TELCOM 2110 62 31