Game Theory Applications for Content Delivery Networks

Size: px
Start display at page:

Download "Game Theory Applications for Content Delivery Networks"

Transcription

1 Game Theory Applications for Content Delivery Networks Chaki Ng, Harvard University [ chaki@eecs.harvard.edu ] Version Introduction As Internet topology and user base grow by the minutes, services delivered on the Internet are competing more fiercely against each other for finite network and computing resources. High availability and responsiveness of services are keys to business sites, which may lose valuable customers otherwise. Classic examples of service disasters include Victoria s Secrets and the NASA Pathfinder mission web casts, during which users could not access to the sites or experienced infinite delays. These were caused by the so-called Flash Crowds [4], which are large numbers of users trying to access the same web site simultaneously and thus demanding site resources exponentially. The issue further complicates beyond the site, as traffic from such flash crowds affect the rest of the Internet. Content delivery networks, CDN, are one of the key service management infrastructures that relieve such service degradation. CDN helps its customers, the content providers, run business sites by allowing them to replicate frequently requested objects such as images and videos on CDN cache servers around the globe. This has a few advantages: ) server load is reduced because requests are served by multiple servers; 2) as servers are located across different physical locations and services clients at different locations, network traffic will spread out; and 3) since users will be served by servers located closest to them, the latency the user experience should be reduced. For more background on how CDN works please refer to [5]. We are interested in analyzing several basic technical and business aspects of a Content Delivery Network (CDN) from a game theoretic perspective. CDN is a hot new industry with lots of open and complicated issues. The industry foresees CDNs to further grow into a multi-billion dollar industry, and new technologies that make CDNs better continue to evolve. It is our hope that by starting to study some basic game theoretic issues for CDNs through this paper, we will gain further inspiration for advanced game theory applications in the domain. The three issues we will address are:

2 . How does CDN design its network of cache servers such that network traffic is minimized? 2. For high demand, live streaming events, how can we improve performance in real-time? 3. Are specific Quality of Service agreements advantageous for CDN customers? CDN Basics First we will introduce some terminology. There are three main parties involved in this paper. A content provider, or customer, is one that operates a website that is open to public access (e.g. CNN.com). A CDN is a company that operates CDN services for content providers. The CDN uses its caching servers located at different geographical locations to provide the services. A client is a generic person accessing the content provider website directly from anywhere on the planet and does not involve directly with the CDN the fact that content providers use CDN services is transparent to the client. Clients () All client requests first arrive at CNN site (2) CNN returns basic docs (e.g. cnn/index.html) CNN Get new objects in cache (3) Selected embedded objects (e.g. graphics, movies) will be served by Akamai bos (5) Objects served by servers that are closest and not overloaded sf (4) Forward requests Selection Algorithms uk Servers around the globe Akamai Figure : CDN Workflow Mechanism Figure shows how a typical CDN works. We use CNN.com as the content provider and Akamai as the CDN. First, clients from around the globe all access CNN.com the same way, by using in their browser. Second, when CNN.com receives such a request, its web server will serve the content that it decides to serve internally. These include low bandwidth and 2

3 light items such as the skeleton homepage itself and small graphics. Third, for items that are frequently requested and high bandwidth, CNN.com would automatically forward the client requests to its CDN provider, Akamai. Akamai would then know about the objects CNN.com wants it to serve and the locations of the clients. It needs to determine, for each client, which caching server should deliver the objects. This is done by Akamai s proprietary algorithms. Typically, for a specific client located say in Boston, Akamai will choose from its caching servers in the Boston area and select the one that is not currently overloaded. On first request the selected caching server does not yet have the objects cached in its space, and thus will download the objects directly from CNN.com. This is only done once, and for any subsequent requests the server would simply serve the objects from its own local space. CDN Network Design In this section we will attempt to address question : How does CDN design its network of cache servers such that network traffic is minimized? Because the purpose of the CDN is to improve performance for business sites, it is important that the CDN network of caching servers has good performance. First of all we have to distinguish between what a CDN does and does not have control over to improve performance. A CDN does not have control over the Internet traffic. The Internet is a massive collection of different networks and domains, each owned by different administrative entities (e.g. universities, corporations, government, etc). Such decentralized setup is the main reason why the Internet has blossomed. Traffic can freely flow from one place to another based on individual (the routers that send data round) goals, which are mainly looking for the lowest traffic and latency at all times. A CDN has control over the caching servers and how they are connected among each other and to the business sites. Essentially a CDN is building a global scale network of thousands of caching servers. As such network is built on top of another existing infrastructure, the Internet, and routing of data only exists among these caching servers, it is called an overlay network. By adding an extra abstract layer under its full control, the CDN can derive proprietary algorithms and mange the infrastructure without having bottlenecks from other network domains. In other words, the CDN can choose the best underlying Internet paths that one caching server has to go through to communicate with another server or the customer website (for requested caching objects) and avoid the congested routes. We shall try to model such traffic design as games and see what can be applied in designing CDN network. Our work is based heavily on [2]. 3

4 We will start with some basics on how selfish agents affect Internet traffic. Suppose there is an origin node, FROM, and a destination node, TO. Data traffic goes from the origin to the destination. There are two paths that data can choose from, UPPER and LOWER. Each path has some latency factor, which means that data flowing through a path will be delayed by that factor. For UPPER the latency factor is X, meaning that it is equal to the amount of data flowing through this path. For LOWER the latency is. To calculate latency of a path, we multiply the traffic by the latency factor. Figure 2 shows a specific example. Total data traffic flowing is. Exactly half of the traffic chooses UPPER, and the other half chooses LOWER. The latency for UPPER is ½ * X = ½ * ½ = ¼. The latency for LOWER is ½ * = ½. Total latency between FROM and TO is thus ¼ + ½ = ¾. upper X Flow = /2 From To lower Flow = /2 Figure 2 Is this example realistic? No. Because the two paths yield different latencies, data packets will all prefer to travel to the lower latency path (UPPER). We can verify this with a simple 2 player game as shown in Figure 3. Packet B Upper Lower /2 Upper /2 /2 Packet A /2 Lower Figure 3 4

5 We make things simple by assuming only two packets make up the flow and that they both contribute to ½ of the total flow (equals one). The payoffs in this game are the latency factor that the packets know. Note that we use latency factor, and not latency, because packets do not know beforehand how much traffic will also decide to travel on the same paths. Latencies are calculated after everyone makes its move. Hence, the payoffs are either ½ if a packet chooses UPPER or if it chooses LOWER. Since lower payoffs are better in this game, the Nash Equilibrium of the game is [UPPER, UPPER]. The resulting path choices are depicted in Figure 4: upper X Flow = From To lower Flow = Figure 4 Because both packets choose UPPER, we have a flow of ½ + ½ = for UPPER. Because there is no traffic for LOWER, total latency equals flow for upper + flow for lower = * (remember X means the same as the flow value itself) + * =. We can see that in Nash Equilibrium the total latency is not optimal! We could have done better as in Figure 2 by having packets split paths and obtain a lower latency of ¾ instead. This result is not unexpected since we have learned that Nash Equilibrium is not necessarily optimal. Thus far, we have used a simple latency function which is linear (X and ). In [2] they have shown results for non-linear latency functions and found the Nash latency would be at worst a factor of two of the optimal path combinations. In their paper, it was suggested that we simply double the bandwidth. However, doubling is easier said than done. Recall that the Internet is a network of networks. It is much easier to double the major backbones such as those owned by the major ISP companies like AT&T (analogy: the Interstate highway system) than the lower level links (analogy: local streets). For example, lots of users these days are connected with cable modem or DSL modem, which cannot be doubled in bandwidth due to technological constraints. Hence, doubling would only work on some paths, but not all paths in the CDN network. Therefore, we still would constantly experience latency paths. The main reason is that even if we double the latency, we 5

6 have not solved the selfishness issue. Selfish packets still are looking for the low latency paths, which are the ones that we just doubled. We are now ready to see how CDN can use the above behaviors to design its network. One intuition one may have is to allow more paths with low latency factors be connected together. Suppose we have the paths shown in Figure 5 connecting two CDN caching servers physically on the Internet. Each of the two paths joining FROM and TO includes two sub-paths that have latency and X. In other words the latency of the two are equal (they are just reversed in order). Since CDN has control over how FROM and TO are connected, it will likely want to improve the latency by allowing traffic to go through only the low latency sub-paths. Figure 6 depicts how this is done. CDN establishes a new sub-path (to simplify the example we assume it has a latency of zero) joining the two intermediate nodes, so that traffic from the low latency UPPER path can go through it and join the low latency LOWER path. Thus the latency factors are X--X. Because this path is preferred to other paths, all traffic flow through this path. Flow = /2 X From To X Flow = /2 Figure 5 6

7 Flow = X From To X Figure 6 We can calculate the total latency in both cases: Total latency of Figure 5 = latency of UPPER (with ½ of the flow) + latency of LOWER (with ½ of the flow) = (½ * X + ½ * ) + (½ * + ½ *X) =.5 Total latency of Figure 6 (only one path, flow = ) = latency of sub-path UPPER (X) + latency of new path () + latency of sub-path LOWER (X) = * X + * + * X = * + * = 2 The supposedly new design of the CDN yields higher total latency!? This puzzle is the classic Braess s dilemma ([7]). What this result shows is that, when designing the network, a CDN does not want to join only the low latency paths together. All traffic will go through those optimal paths, and the overall latency is higher than more balanced paths like those in Figure 5. In summary, the CDN should evaluate all possible path combinations, and find path combinations such that the total latency is minimized. For example, if the CDN is given possible paths in Figure 6, it should attempt to take the zero latency path out, so that it can attain the optimal solution as in Figure 5. In essence, standard way of thinking about optimization does not apply in this latency problem. One must consider the independent actions of all agents involve and provide a specific set of rules that the agents must follow. 7

8 QoS Guarantees or Not? In this section, we will analyze whether specific Quality of Service agreements are advantageous for CDN customers. Typically, like many other Internet services, CDNs sign fixed fee contracts with customers with no specific service guarantees. The most common service guarantee on those contracts would be about up time (e.g. 99% up time ), which means that the service will be unavailable less than % anytime. Even if 99% of the requests by the customer are served by the CDN, the customer has no control over other parameters. For example, the CDN does not guarantee how long it will take for a client to download all the objects. Content providers like Amazon would want to make sure that its web pages are fully downloaded to its clients as quickly as possible (the industry often talks about a 7-second rule), because if it takes too long the clients can easily go to Amazon s competitors like Barnes and Noble to buy the same products. Similarly, for important client requests (e.g. someone is on the online shopping checkout page), a customer would not want the CDN to process these with overloaded caching servers. A new trend is to change the terms of the contracts, such that there are some guarantees of specific services. With the above two examples, the agreement may include QoS terms like 9% of all requests must be served within seconds and % of the requests are served by caching servers with load no more than 8%. We will model this as a game and use Forward Induction to solve it. First, we will ignore QoS for now and frame the standard game that CDN and content provider play. For content provider, there are two choices: INCREASE LOAD, or NORMAL. NORMAL means content provider would use the service as-is, while INCREASE LOAD means the content provider will try to increase routing requests to the CDN. At NORMAL content provider will be indifferent with the service CDN provides (in other words, CDN can deliver NORMAL load services reasonably well). At INCREASE LOAD the content provider is hoping that by using more of the CDN s services its customers will be happier. For CDN, it has two choices: PRIORITY or NO. For most content providers, CDN would treat them the same and with NO priority. If it decides to treat a particular content provider with PRIORITY, then the services delivered for that content provider must be satisfactory. Why would a CDN assign PRIORITY to some content providers? Perhaps those content providers are the biggest clients of the CDN or they received bad services from CDN recently and it is trying to keep the relationships. 8

9 Increase Load Content Provider Normal 3 Priority - CDN - No Figure 7 Figure 7 depicts the choices and payoffs. When CDN treat a content provider with NO priority, and content provider uses CDN services with NORMAL load, both have zero payoffs (services are OK). If CDN treats this NORMAL load with PRIORITY, CDN would have a - payoff. This is an opportunity cost for CDN. Since content provider is just using the normal load, CDN could have waived priority and could provide high priority services elsewhere. Now we will see what happens when content provider increases load. If CDN does not make it a priority, then the content provider would have negative payoff of -. Its clients would receive bad services, and it will end up losing revenue and clients. There is a zero payoff for the CDN, since it has not done anything for the situation and is not required to do so. If CDN makes it a priority, then the content provider would receive a high payoff of 3. It invites more clients to use its services via CDN, and the services are of quality. CDN makes the right choice and it will receive a payoff of (the content provider is happy and would likely renew their agreement in the future). 9

10 Increase Load Content Provider Normal 2 - Priority 2 CDN -2 - No Figure 8 Now we will add in the QoS choice. To active QoS agreement, content provider must first pay CDN a fee, similar to buying insurance, for the guarantee. The cost for content provider would be and it adds directly to CDN s payoffs. Figure 8 shows the new payoffs. The deal of the QoS normally works like this. If CDN meets the QoS item, then it will receive an additional payoff of as bonus. The content provider serves its customers well, makes an additional profit of 3 because of it. Its net profit is hence 2 after paying the bonus to the CDN. If the QoS is not met, the CDN will have to refund the QoS fee plus pay a penalty of 2 to content provider. Thus, the payoff of this game with QoS is more complicated than some examples we have seen in examples in class, where the option available would only have a constant payoff effect (+ on a player, and - on the other) on each outcome combinations. Increase Load Content Provider Normal 4 - Priority 3 CDN - No -2 Figure 9

11 The full game is modeled in Figure. We will use Forward Deduction to find the solution. The procedures are: () Remove NORMAL under QoS branch for content provider. It could have stay with no QoS and receive a payoff of rather than -. (2) CDN then can eliminate NO since it will choose PRIORITY with a higher payoff of 3. (3) NORMAL and No QoS for content provider is strongly dominated by QoS INCREASE LOAD ( vs. 4). (4) CDN will choose PRIORITY over NO ( vs. ). (5) We are left with two choices, both PRIORITY, INCREASE LOAD. Content provider will go for QoS guarantee and get payoff of 4. Content Provider QoS Quarantee No (5) Increase Load Content Provider Normal Increase Load Content Provider Normal () Priority Priority - 3 CDN CDN - - No No -2 (4) (3) (2) Figure The result of this game is not surprising. Even after just steps () and (2), we are basically left with the game of the left hand size (with no QoS) and [PRIORITY, INCREASE LOAD] on the right. The payoffs of [PRIORITY, INCREASE] is simply better than any of those in no QoS. This choice

12 is basically a win-win situation for both parties: CDN gets paid to deliver high quality services, which it is capable of doing, and get rewarded with bonuses that other non-qos customers would not pay. Content provider can now offload most of its operations to CDN and focuses on its core business (recruit and retain clients, generate revenue). Content provider also knows that CDN will deliver the highest possible services it can deliver, and rewards CDN with profit sharing (the bonus). Streaming Media Multicasting Sometimes content providers hire CDNs for special web cast events. These are live streaming media broadcast over the Internet, such as a major concert event. The amount of simultaneous requests for such web casts is typically extremely high (for popular events there will be millions of simultaneous requests). Also, streaming media files are much bigger than typical graphics and media file, and because users are viewing the files in real-time, it is necessary that the streaming not be interrupted. Achieving high quality and high volume web casts is not easy. The CDNs uses a solution called Multicasting to address this problem. Basically, the idea is that many requests come from the same paths (group requests in regions, say Cambridge / Boston / Worcester / Springfield, think of each region as a number of clients connecting via the same ISP, hence sharing paths). Sending data on the Internet does not have marginal cost benefits sending the data on a path to one user costs the same as sending to multiple users. Hence, it is more economical to send the streaming files one time along those paths and get as many users as possible to receive them. Interesting work by [6] demonstrates mechanisms for implementing multicasting with incentive compatible mechanism. It uses similar settings as cooperative game theory and the marginal contribution concepts. We shall attempt to apply it to an example and see how it may work for a CDN. We will focus on one of the caching servers of the CDN, say the one in Boston serving clients in the area for a live web cast of a World Cup game. Suppose there are four different destinations where all the clients reside. Figure shows the setup. There are costs associated with each path from the caching server to the destinations (the numbers next to the paths). Each client has a willingness-to-pay value to get the web cast (e.g. {3, 8} means that two clients share the same path, and they have values of 3 and 8 respectively). 2

13 Caching Server {3, 8} {4, 38, 25, } {3, 25} {3, 25, 5} Figure First of all, the caching server has to decide which destinations to send the web cast to, based on net profits. If a destination is accepted, then all clients at that destination are accepted as well (since there is no additional cost). The destinations that have combined client values higher than the path costs would be selected, and we shall apply this rule to each path, and each subset of the whole tree. Checking the lower four paths we can see that they each have combined client values higher than costs. For example, 3+8>2, >4. We then consider the sub trees branching out from the caching server node. Both sides have overall values higher than costs (e.g. the left size client values are higher than = 9). We can conclude that the caching server will serve all of the clients in this example. Second, we need to find out the total value, or pie, of this tree. The total cost of the multicast tree is the same of the cost of each edge. That equals to = 96. The total value = = 276. The pie created is value cost = = 8. We will call this V(N). Third, we calculate the marginal contributions of each client. This is done by finding V(N) V(N\i), where V(N\i) is the total value without client i. To calculate V(N\) for client, we first need to calculate the total costs without it. In this case, the destination belongs will not be served, because we now only have one client with a value of 8 < the cost of the path = 2. Therefore, the total value without = (276 38) (96 2) = 62. Since V(N) = 8, the marginal contribution of client = 8-62 = 8. Similarly, the marginal contributions for the rest of the clients are {8, 4, 38, 25,, 9, 9, 9, 9, 5}. 3

14 [6] suggest payment schemes that would allow this to be implemented. Payment for each client would be its individual value minus marginal contribution. Again for client its payment would be 3 8 = 2. The revenue for the tree would be the total costs of the paths. With this scheme, CDN can serve content providers for web casts with a fair pricing system. In practice, actual clients are not aware of the multicasting structure, and their values are determined by software agents installed by the content providers. Conclusion We have demonstrated three basic CDN issues that can be analyzed using game theoretic methods. The biggest issue in computer science is that all things are assumed to be cooperative (or non-selfish). Because of the Internet things have changed, and one cannot assume other resources/programs would be cooperative. The infusion of game theory with computer science seems very attractive to create new models of computation and analytical thinking. Content delivery networks are one of the most widely used applications on the Internet and they keep it from congestion. We hope to further develop some of these methods and bring them to the realworld. Reference [] Krishnamurthy, Balachander, Craig Wills, and Yin Zhang. On the Use and Performance of Content Distribution Networks. ACM Internet Measurement Workshop, 2. [2] Roughgarden, Tim and Eva Tardos. How Bad is Selfish Routing? FOCS 2. [3] Intel ebusiness Case Studies on Akamai. ( [4] Ratul Mahajan, Steven M. Bellovin, Sally Floyd, John Ioannidisand Vern Paxson, and Scott Shenker. Controlling high bandwidth aggregates in the network. Technical report, (draft), February 2. ( [5] Rabinovich, Michael, and Oliver Spatscheck. Web Caching and Replication. 2. [6] Feigenbaum, J., C. Papadimitriou and S. Shenker. Sharing the Cost of Multicast Transmissions. Journal of Computer and System Sciences 63 (2), pp [7] Braess, D. Uber ein paradoxon der verkehrsplanung. Unternehmensforschung, 2: ,

The CDN Internetworking Auction

The CDN Internetworking Auction Market Models for Content Delivery Networks Introduction Chaki Ng, Harvard University [chaki@eecs.harvard.edu] Version 5.16.2002 As Internet topology and user base grow by the minutes, services delivered

More information

Distributed Systems. 23. Content Delivery Networks (CDN) Paul Krzyzanowski. Rutgers University. Fall 2015

Distributed Systems. 23. Content Delivery Networks (CDN) Paul Krzyzanowski. Rutgers University. Fall 2015 Distributed Systems 23. Content Delivery Networks (CDN) Paul Krzyzanowski Rutgers University Fall 2015 November 17, 2015 2014-2015 Paul Krzyzanowski 1 Motivation Serving web content from one location presents

More information

Data Center Content Delivery Network

Data Center Content Delivery Network BM 465E Distributed Systems Lecture 4 Networking (cont.) Mehmet Demirci Today Overlay networks Data centers Content delivery networks Overlay Network A virtual network built on top of another network Overlay

More information

How To Understand The Power Of A Content Delivery Network (Cdn)

How To Understand The Power Of A Content Delivery Network (Cdn) Overview 5-44 5-44 Computer Networking 5-64 Lecture 8: Delivering Content Content Delivery Networks Peter Steenkiste Fall 04 www.cs.cmu.edu/~prs/5-44-f4 Web Consistent hashing Peer-to-peer CDN Motivation

More information

Distributed Systems. 25. Content Delivery Networks (CDN) 2014 Paul Krzyzanowski. Rutgers University. Fall 2014

Distributed Systems. 25. Content Delivery Networks (CDN) 2014 Paul Krzyzanowski. Rutgers University. Fall 2014 Distributed Systems 25. Content Delivery Networks (CDN) Paul Krzyzanowski Rutgers University Fall 2014 November 16, 2014 2014 Paul Krzyzanowski 1 Motivation Serving web content from one location presents

More information

Web Caching and CDNs. Aditya Akella

Web Caching and CDNs. Aditya Akella Web Caching and CDNs Aditya Akella 1 Where can bottlenecks occur? First mile: client to its ISPs Last mile: server to its ISP Server: compute/memory limitations ISP interconnections/peerings: congestion

More information

BRAESS-LIKE PARADOXES FOR NON-COOPERATIVE DYNAMIC LOAD BALANCING IN DISTRIBUTED COMPUTER SYSTEMS

BRAESS-LIKE PARADOXES FOR NON-COOPERATIVE DYNAMIC LOAD BALANCING IN DISTRIBUTED COMPUTER SYSTEMS GESJ: Computer Science and Telecommunications 21 No.3(26) BRAESS-LIKE PARADOXES FOR NON-COOPERATIVE DYNAMIC LOAD BALANCING IN DISTRIBUTED COMPUTER SYSTEMS Said Fathy El-Zoghdy Department of Computer Science,

More information

Measuring the Web: Part I - - Content Delivery Networks. Prof. Anja Feldmann, Ph.D. Dr. Ramin Khalili Georgios Smaragdakis, PhD

Measuring the Web: Part I - - Content Delivery Networks. Prof. Anja Feldmann, Ph.D. Dr. Ramin Khalili Georgios Smaragdakis, PhD Measuring the Web: Part I - - Content Delivery Networks Prof. Anja Feldmann, Ph.D. Dr. Ramin Khalili Georgios Smaragdakis, PhD Acknowledgement Material presented in these slides is borrowed from presentajons

More information

DATA COMMUNICATOIN NETWORKING

DATA COMMUNICATOIN NETWORKING DATA COMMUNICATOIN NETWORKING Instructor: Ouldooz Baghban Karimi Course Book: Computer Networking, A Top-Down Approach, Kurose, Ross Slides: - Course book Slides - Slides from Princeton University COS461

More information

Azure Media Service Cloud Video Delivery KILROY HUGHES MICROSOFT AZURE MEDIA 2015.08.20

Azure Media Service Cloud Video Delivery KILROY HUGHES MICROSOFT AZURE MEDIA 2015.08.20 Azure Media Service Cloud Video Delivery KILROY HUGHES MICROSOFT AZURE MEDIA 2015.08.20 Azure Cloud Topology Public cloud providers such as Amazon Web Service, Google, IBM, Rackspace, etc. have similar

More information

Demystifying CDNs: Building Video Enabled Networks That Maintain Your Network Integrity And Meet User Demand

Demystifying CDNs: Building Video Enabled Networks That Maintain Your Network Integrity And Meet User Demand An Enterprise Video Communications Industry Lite Paper Demystifying CDNs: Building Video Enabled Networks That Maintain Your Network Integrity And Meet User Demand Overview In recent years video has taken

More information

Ensuring Real-Time Traffic Quality

Ensuring Real-Time Traffic Quality Ensuring Real-Time Traffic Quality Summary Voice and video calls are traffic that must arrive without delay at the receiving end for its content to be intelligible. This real-time traffic is different

More information

Peer-to-Peer Networks. Chapter 6: P2P Content Distribution

Peer-to-Peer Networks. Chapter 6: P2P Content Distribution Peer-to-Peer Networks Chapter 6: P2P Content Distribution Chapter Outline Content distribution overview Why P2P content distribution? Network coding Peer-to-peer multicast Kangasharju: Peer-to-Peer Networks

More information

Content Delivery Networks. Shaxun Chen April 21, 2009

Content Delivery Networks. Shaxun Chen April 21, 2009 Content Delivery Networks Shaxun Chen April 21, 2009 Outline Introduction to CDN An Industry Example: Akamai A Research Example: CDN over Mobile Networks Conclusion Outline Introduction to CDN An Industry

More information

Octoshape s Multicast Technology Suite:

Octoshape s Multicast Technology Suite: : The Next-Gen CDN Alternative for Large-Scale, Cost-Optimized, Global HD Streaming HQ: +45 8833 4680 USA: +1 770 578 1686 Asia: +65 81125330 www.octoshape.com Table of Contents Core Transport...4 Making

More information

DNS, CDNs Weds March 17 2010 Lecture 13. What is the relationship between a domain name (e.g., youtube.com) and an IP address?

DNS, CDNs Weds March 17 2010 Lecture 13. What is the relationship between a domain name (e.g., youtube.com) and an IP address? DNS, CDNs Weds March 17 2010 Lecture 13 DNS What is the relationship between a domain name (e.g., youtube.com) and an IP address? DNS is the system that determines this mapping. Basic idea: You contact

More information

Computer Network. Interconnected collection of autonomous computers that are able to exchange information

Computer Network. Interconnected collection of autonomous computers that are able to exchange information Introduction Computer Network. Interconnected collection of autonomous computers that are able to exchange information No master/slave relationship between the computers in the network Data Communications.

More information

Intelligent Content Delivery Network (CDN) The New Generation of High-Quality Network

Intelligent Content Delivery Network (CDN) The New Generation of High-Quality Network White paper Intelligent Content Delivery Network (CDN) The New Generation of High-Quality Network July 2001 Executive Summary Rich media content like audio and video streaming over the Internet is becoming

More information

Scalable Data Collection for Internet-based Digital Government Applications

Scalable Data Collection for Internet-based Digital Government Applications Scalable Data Collection for Internet-based Digital Government Applications [ Appeared in Proceedings of the 1st National Conference on Digital Government Research, 2001 ] W. C. Cheng C.-F. Chou L. Golubchik

More information

Distributed Systems. 24. Content Delivery Networks (CDN) 2013 Paul Krzyzanowski. Rutgers University. Fall 2013

Distributed Systems. 24. Content Delivery Networks (CDN) 2013 Paul Krzyzanowski. Rutgers University. Fall 2013 Distributed Systems 24. Content Delivery Networks (CDN) Paul Krzyzanowski Rutgers University Fall 2013 November 27, 2013 2013 Paul Krzyzanowski 1 Motivation Serving web content from one location presents

More information

Indirection. science can be solved by adding another level of indirection" -- Butler Lampson. "Every problem in computer

Indirection. science can be solved by adding another level of indirection -- Butler Lampson. Every problem in computer Indirection Indirection: rather than reference an entity directly, reference it ( indirectly ) via another entity, which in turn can or will access the original entity A x B "Every problem in computer

More information

Teridion. Rethinking Network Performance. The Internet. Lightning Fast. Technical White Paper July, 2015 www.teridion.com

Teridion. Rethinking Network Performance. The Internet. Lightning Fast. Technical White Paper July, 2015 www.teridion.com Teridion The Internet. Lightning Fast. Rethinking Network Performance Technical White Paper July, 2015 www.teridion.com Executive summary Online services face the growing dual challenge of supporting many

More information

Internet Content Distribution

Internet Content Distribution Internet Content Distribution Chapter 2: Server-Side Techniques (TUD Student Use Only) Chapter Outline Server-side techniques for content distribution Goals Mirrors Server farms Surrogates DNS load balancing

More information

PQoS Parameterized Quality of Service. White Paper

PQoS Parameterized Quality of Service. White Paper P Parameterized Quality of Service White Paper Abstract The essential promise of MoCA no new wires, no service calls and no interference with other networks or consumer electronic devices remains intact

More information

HPAM: Hybrid Protocol for Application Level Multicast. Yeo Chai Kiat

HPAM: Hybrid Protocol for Application Level Multicast. Yeo Chai Kiat HPAM: Hybrid Protocol for Application Level Multicast Yeo Chai Kiat Scope 1. Introduction 2. Hybrid Protocol for Application Level Multicast (HPAM) 3. Features of HPAM 4. Conclusion 1. Introduction Video

More information

The old Internet. Software in the Network: Outline. Traditional Design. 1) Basic Caching. The Arrival of Software (in the network)

The old Internet. Software in the Network: Outline. Traditional Design. 1) Basic Caching. The Arrival of Software (in the network) The old Software in the Network: What Happened and Where to Go Prof. Eric A. Brewer UC Berkeley Inktomi Corporation Local networks with local names and switches IP creates global namespace and links the

More information

How To Provide Qos Based Routing In The Internet

How To Provide Qos Based Routing In The Internet CHAPTER 2 QoS ROUTING AND ITS ROLE IN QOS PARADIGM 22 QoS ROUTING AND ITS ROLE IN QOS PARADIGM 2.1 INTRODUCTION As the main emphasis of the present research work is on achieving QoS in routing, hence this

More information

AKAMAI WHITE PAPER. Delivering Dynamic Web Content in Cloud Computing Applications: HTTP resource download performance modelling

AKAMAI WHITE PAPER. Delivering Dynamic Web Content in Cloud Computing Applications: HTTP resource download performance modelling AKAMAI WHITE PAPER Delivering Dynamic Web Content in Cloud Computing Applications: HTTP resource download performance modelling Delivering Dynamic Web Content in Cloud Computing Applications 1 Overview

More information

Distributed Systems 19. Content Delivery Networks (CDN) Paul Krzyzanowski pxk@cs.rutgers.edu

Distributed Systems 19. Content Delivery Networks (CDN) Paul Krzyzanowski pxk@cs.rutgers.edu Distributed Systems 19. Content Delivery Networks (CDN) Paul Krzyzanowski pxk@cs.rutgers.edu 1 Motivation Serving web content from one location presents problems Scalability Reliability Performance Flash

More information

Stability of QOS. Avinash Varadarajan, Subhransu Maji {avinash,smaji}@cs.berkeley.edu

Stability of QOS. Avinash Varadarajan, Subhransu Maji {avinash,smaji}@cs.berkeley.edu Stability of QOS Avinash Varadarajan, Subhransu Maji {avinash,smaji}@cs.berkeley.edu Abstract Given a choice between two services, rest of the things being equal, it is natural to prefer the one with more

More information

Internet Content Distribution

Internet Content Distribution Internet Content Distribution Chapter 4: Content Distribution Networks (TUD Student Use Only) Chapter Outline Basics of content distribution networks (CDN) Why CDN? How do they work? Client redirection

More information

Dynamic Load Balancing and Node Migration in a Continuous Media Network

Dynamic Load Balancing and Node Migration in a Continuous Media Network Dynamic Load Balancing and Node Migration in a Continuous Media Network Anthony J. Howe Supervisor: Dr. Mantis Cheng University of Victoria Draft: April 9, 2001 Abstract This report examines current technologies

More information

Flash Crowds & Denial of Service Attacks

Flash Crowds & Denial of Service Attacks Flash Crowds & Denial of Service Attacks Characterization and Implications for CDNs and Web sites Jaeyeon Jung MIT Laboratory for Computer Science Balachander Krishnamurthy and Michael Rabinovich AT&T

More information

Lecture 3: Scaling by Load Balancing 1. Comments on reviews i. 2. Topic 1: Scalability a. QUESTION: What are problems? i. These papers look at

Lecture 3: Scaling by Load Balancing 1. Comments on reviews i. 2. Topic 1: Scalability a. QUESTION: What are problems? i. These papers look at Lecture 3: Scaling by Load Balancing 1. Comments on reviews i. 2. Topic 1: Scalability a. QUESTION: What are problems? i. These papers look at distributing load b. QUESTION: What is the context? i. How

More information

Content Distribution over IP: Developments and Challenges

Content Distribution over IP: Developments and Challenges Content Distribution over IP: Developments and Challenges Adrian Popescu, Blekinge Inst of Technology, Sweden Markus Fiedler, Blekinge Inst of Technology, Sweden Demetres D. Kouvatsos, University of Bradford,

More information

Content Distribu-on Networks (CDNs)

Content Distribu-on Networks (CDNs) Content Distribu-on Networks (CDNs) Jennifer Rexford COS 461: Computer Networks Lectures: MW 10-10:0am in Architecture N101 hjp://www.cs.princeton.edu/courses/archive/spr12/cos461/ Second Half of the Course

More information

1. Comments on reviews a. Need to avoid just summarizing web page asks you for:

1. Comments on reviews a. Need to avoid just summarizing web page asks you for: 1. Comments on reviews a. Need to avoid just summarizing web page asks you for: i. A one or two sentence summary of the paper ii. A description of the problem they were trying to solve iii. A summary of

More information

Guiding Web Proxy and Server Placement for High Performance Internet Content Delivery 1

Guiding Web Proxy and Server Placement for High Performance Internet Content Delivery 1 Guiding Web Proxy and Server Placement for High Performance Internet Content Delivery 1 Peter Triantafillou (contact author) Department of Computer Engineering and Informatics, University of Patras Rio

More information

Testing & Assuring Mobile End User Experience Before Production. Neotys

Testing & Assuring Mobile End User Experience Before Production. Neotys Testing & Assuring Mobile End User Experience Before Production Neotys Agenda Introduction The challenges Best practices NeoLoad mobile capabilities Mobile devices are used more and more At Home In 2014,

More information

How To Test Performance Of A Cdn Server

How To Test Performance Of A Cdn Server On the Use and Performance of Content Distribution Networks Yin Zhang Joint work with Balachander Krishnamurthy and Craig Wills AT&T Labs Research, WPI {yzhang,bala}@research.att.com, cew@cs.wpi.edu ACM

More information

Octoshape. Introducing. a new technology for large-scale streaming over the Internet. Scale and cost problems are the spoilers

Octoshape. Introducing. a new technology for large-scale streaming over the Internet. Scale and cost problems are the spoilers Octoshape Introducing a new technology for large-scale streaming over the Internet Stephen Alstrup and Theis Rauhe Octoshape The popularity of live streaming over the Internet is growing. The number of

More information

The Requirement for a New Type of Cloud Based CDN

The Requirement for a New Type of Cloud Based CDN The Requirement for a New Type of Cloud Based CDN Executive Summary The growing use of SaaS-based applications has highlighted some of the fundamental weaknesses of the Internet that significantly impact

More information

End-to-end Quality of Service. Robert Rudin, Chaki Ng, Clifford Kahn 5/16/2006

End-to-end Quality of Service. Robert Rudin, Chaki Ng, Clifford Kahn 5/16/2006 End-to-end Quality of Service Robert Rudin, Chaki Ng, Clifford Kahn 5/16/2006 1 Conclusions Major barrier to service-level interconnection is coordination Need a coordinator an overlay Network Neutrality:

More information

Region 10 Videoconference Network (R10VN)

Region 10 Videoconference Network (R10VN) Region 10 Videoconference Network (R10VN) Network Considerations & Guidelines 1 What Causes A Poor Video Call? There are several factors that can affect a videoconference call. The two biggest culprits

More information

Traffic delivery evolution in the Internet ENOG 4 Moscow 23 rd October 2012

Traffic delivery evolution in the Internet ENOG 4 Moscow 23 rd October 2012 Traffic delivery evolution in the Internet ENOG 4 Moscow 23 rd October 2012 January 29th, 2008 Christian Kaufmann Director Network Architecture Akamai Technologies, Inc. way-back machine Web 1998 way-back

More information

How To Improve Performance On A Ccdn (Dns)

How To Improve Performance On A Ccdn (Dns) Enhanced Content Delivery Network to Improve the QoE 1 Sachendra Singh Solanky, 2 Sandra Brigit Johnson, 3 Vakkalagadda Eswar Praphul 1 M.Tech Student SCSE, VIT University Chennai-600048, 2 M.Tech Student

More information

Choosing a Content Delivery Method

Choosing a Content Delivery Method Choosing a Content Delivery Method Executive Summary Cache-based content distribution networks (CDNs) reach very large volumes of highly dispersed end users by duplicating centrally hosted video, audio

More information

Scalable Internet/Scalable Storage. Seif Haridi KTH/SICS

Scalable Internet/Scalable Storage. Seif Haridi KTH/SICS Scalable Internet/Scalable Storage Seif Haridi KTH/SICS Interdisk: The Big Idea 2 Interdisk: The Big Idea I: 3 Interdisk: The Big Idea I: Internet is global data communication 4 Interdisk: The Big Idea

More information

Access the Test Here http://myspeed.visualware.com/index.php

Access the Test Here http://myspeed.visualware.com/index.php VoIP Speed Test Why run the test? Running a VoIP speed test is an effective way to gauge whether your Internet connection is suitable to run a hosted telephone system using VoIP technology. A number of

More information

Introduction to IP v6

Introduction to IP v6 IP v 1-3: defined and replaced Introduction to IP v6 IP v4 - current version; 20 years old IP v5 - streams protocol IP v6 - replacement for IP v4 During developments it was called IPng - Next Generation

More information

The Value of a Content Delivery Network

The Value of a Content Delivery Network September 2010 White Paper The Value of a Content Delivery Network Table of Contents Introduction... 3 Performance... 3 The Second Generation of CDNs... 6 Conclusion... 7 About NTT America... 8 Introduction

More information

How To Understand The Power Of Icdn

How To Understand The Power Of Icdn MobiArch 2014 R-iCDN: an Approach Supporting Flexible Content Routing for ISP-operated CDN Song Ci High Performance Network Lab, Institute of Acoustics, Chinese Academy of Sciences Outline I. Background

More information

CS 40, Lecture 3: Internet economics. Ramesh Johari

CS 40, Lecture 3: Internet economics. Ramesh Johari CS 40, Lecture 3: Internet economics Ramesh Johari Outline Contracts: transit and peer Example 1: Peering and pricing Example 2: Exchanges Example 3: Hot potato routing Example 4: 95 th percentile pricing

More information

QoS issues in Voice over IP

QoS issues in Voice over IP COMP9333 Advance Computer Networks Mini Conference QoS issues in Voice over IP Student ID: 3058224 Student ID: 3043237 Student ID: 3036281 Student ID: 3025715 QoS issues in Voice over IP Abstract: This

More information

Live Streaming with Content Centric Networking

Live Streaming with Content Centric Networking Live Streaming with Content Centric Networking Hongfeng Xu 2,3, Zhen Chen 1,3, Rui Chen 2,3, Junwei Cao 1,3 1 Research Institute of Information Technology 2 Department of Computer Science and Technology

More information

Fundamentals of VoIP Call Quality Monitoring & Troubleshooting. 2014, SolarWinds Worldwide, LLC. All rights reserved. Follow SolarWinds:

Fundamentals of VoIP Call Quality Monitoring & Troubleshooting. 2014, SolarWinds Worldwide, LLC. All rights reserved. Follow SolarWinds: Fundamentals of VoIP Call Quality Monitoring & Troubleshooting 2014, SolarWinds Worldwide, LLC. All rights reserved. Introduction Voice over IP, or VoIP, refers to the delivery of voice and multimedia

More information

Measuring CDN Performance. Hooman Beheshti, VP Technology

Measuring CDN Performance. Hooman Beheshti, VP Technology Measuring CDN Performance Hooman Beheshti, VP Technology Why this matters Performance is one of the main reasons we use a CDN Seems easy to measure, but isn t Performance is an easy way to comparison shop

More information

1.1. Abstract. 1.2. VPN Overview

1.1. Abstract. 1.2. VPN Overview 1.1. Abstract Traditionally organizations have designed their VPN networks using layer 2 WANs that provide emulated leased lines. In the last years a great variety of VPN technologies has appeared, making

More information

SiteCelerate white paper

SiteCelerate white paper SiteCelerate white paper Arahe Solutions SITECELERATE OVERVIEW As enterprises increases their investment in Web applications, Portal and websites and as usage of these applications increase, performance

More information

Controlling the Internet in the era of Software Defined and Virtualized Networks. Fernando Paganini Universidad ORT Uruguay

Controlling the Internet in the era of Software Defined and Virtualized Networks. Fernando Paganini Universidad ORT Uruguay Controlling the Internet in the era of Software Defined and Virtualized Networks Fernando Paganini Universidad ORT Uruguay CDS@20, Caltech 2014 Motivation 1. The Internet grew in its first 30 years with

More information

Lecture Outline. Topology Design: I. Topologies and Reliability. Another Case: Mobile Performance Optimization. Jeremiah Deng.

Lecture Outline. Topology Design: I. Topologies and Reliability. Another Case: Mobile Performance Optimization. Jeremiah Deng. Lecture Outline Topology Design: I 1 Review Jeremiah Deng 2 University of Otago 3 1 / 22 Review Another Case: Mobile Performance Optimization 2 / 22 Topologies and Reliability T. Everts, Rules for mobile

More information

WAN Performance Analysis A Study on the Impact of Windows 7

WAN Performance Analysis A Study on the Impact of Windows 7 A Talari Networks White Paper WAN Performance Analysis A Study on the Impact of Windows 7 Test results demonstrating WAN performance changes due to upgrading to Windows 7 and the network architecture and

More information

Results-Oriented Application Acceleration with FastView Because Every Second Counts Whitepaper

Results-Oriented Application Acceleration with FastView Because Every Second Counts Whitepaper Results-Oriented Application Acceleration with FastView Because Every Second Counts Whitepaper Table of Contents Executive Summary...3 Why Website Performance Matters...3 What Affects Website Performance...5

More information

Optimizing Enterprise Network Bandwidth For Security Applications. Improving Performance Using Antaira s Management Features

Optimizing Enterprise Network Bandwidth For Security Applications. Improving Performance Using Antaira s Management Features Optimizing Enterprise Network Bandwidth For Security Applications Improving Performance Using Antaira s Management Features By: Brian Roth, Product Marketing Engineer April 1, 2014 April 2014 Optimizing

More information

CDN and Traffic-structure

CDN and Traffic-structure CDN and Traffic-structure Outline Basics CDN Traffic Analysis 2 Outline Basics CDN Building Blocks Services Evolution Traffic Analysis 3 A Centralized Web! Slow content must traverse multiple backbones

More information

AN OVERVIEW OF QUALITY OF SERVICE COMPUTER NETWORK

AN OVERVIEW OF QUALITY OF SERVICE COMPUTER NETWORK Abstract AN OVERVIEW OF QUALITY OF SERVICE COMPUTER NETWORK Mrs. Amandeep Kaur, Assistant Professor, Department of Computer Application, Apeejay Institute of Management, Ramamandi, Jalandhar-144001, Punjab,

More information

On the Interaction and Competition among Internet Service Providers

On the Interaction and Competition among Internet Service Providers On the Interaction and Competition among Internet Service Providers Sam C.M. Lee John C.S. Lui + Abstract The current Internet architecture comprises of different privately owned Internet service providers

More information

Study of Flexible Contents Delivery System. With Dynamic Server Deployment

Study of Flexible Contents Delivery System. With Dynamic Server Deployment Study of Flexible Contents Delivery System With Dynamic Server Deployment Yuko KAMIYA Toshihiko SHIMOKAWA and orihiko YOSHIDA Graduate School of Information Science, Kyushu Sangyo University, JAPA Faculty

More information

18: Enhanced Quality of Service

18: Enhanced Quality of Service 18: Enhanced Quality of Service Mark Handley Traditional best-effort queuing behaviour in routers Data transfer: datagrams: individual packets no recognition of flows connectionless: no signalling Forwarding:

More information

Data Center Switch Fabric Competitive Analysis

Data Center Switch Fabric Competitive Analysis Introduction Data Center Switch Fabric Competitive Analysis This paper analyzes Infinetics data center network architecture in the context of the best solutions available today from leading vendors such

More information

Definition. A Historical Example

Definition. A Historical Example Overlay Networks This lecture contains slides created by Ion Stoica (UC Berkeley). Slides used with permission from author. All rights remain with author. Definition Network defines addressing, routing,

More information

Route Control Optimize Multi-homed Connections for Performance, Load and Cost By John Bartlett January 2002

Route Control Optimize Multi-homed Connections for Performance, Load and Cost By John Bartlett January 2002 Route Control Optimize Multi-homed Connections for Performance, Load and Cost By John Bartlett January 2002 The Internet is coming of age, in large part because of its ability to open up markets and to

More information

Troubleshooting Common Issues in VoIP

Troubleshooting Common Issues in VoIP Troubleshooting Common Issues in VoIP 2014, SolarWinds Worldwide, LLC. All rights reserved. Voice over Internet Protocol (VoIP) Introduction Voice over IP, or VoIP, refers to the delivery of voice and

More information

Bandwidth Aggregation, Teaming and Bonding

Bandwidth Aggregation, Teaming and Bonding Bandwidth Aggregation, Teaming and Bonding The increased use of Internet sharing combined with graphically rich web sites and multimedia applications have created a virtually insatiable demand for Internet

More information

March 2010 Webcasting: Dealing with significant audiences behind the corporate firewall

March 2010 Webcasting: Dealing with significant audiences behind the corporate firewall March 2010 Webcasting: Dealing with significant audiences behind the corporate firewall Ed Van Petten CIO / Vice President, Network Operations ON24, Inc. Introduction Webcasts sometimes involve significant

More information

Network Formation and Routing by Strategic Agents using Local Contracts

Network Formation and Routing by Strategic Agents using Local Contracts Network Formation and Routing by Strategic Agents using Local Contracts Elliot Anshelevich 1 and Gordon Wilfong 2 1 Department of Computer Science, Rensselaer Polytechnic Institute, Troy, NY. 2 Bell Labs,

More information

Cost Effective Deployment of VoIP Recording

Cost Effective Deployment of VoIP Recording Cost Effective Deployment of VoIP Recording Purpose This white paper discusses and explains recording of Voice over IP (VoIP) telephony traffic. How can a company deploy VoIP recording with ease and at

More information

The low utilization and high cost of data networks

The low utilization and high cost of data networks The low utilization and high cost of data networks Andrew Odlyzko AT&T Labs - Research amo@research.att.com 1. Introduction The rapid growth of packet data networks is usually ascribed to their lower costs

More information

The Quality of Internet Service: AT&T s Global IP Network Performance Measurements

The Quality of Internet Service: AT&T s Global IP Network Performance Measurements The Quality of Internet Service: AT&T s Global IP Network Performance Measurements In today's economy, corporations need to make the most of opportunities made possible by the Internet, while managing

More information

Challenges in Combining Overlay Networks and Cooperative Caching

Challenges in Combining Overlay Networks and Cooperative Caching Challenges in Combining Overlay Networks and Cooperative Caching Frank T. Johnsen, Trude Hafsøe, Thomas Plagemann and Vera Goebel Dept. of Informatics, University of Oslo, Norway Research Report No 0,

More information

The Role and uses of Peer-to-Peer in file-sharing. Computer Communication & Distributed Systems EDA 390

The Role and uses of Peer-to-Peer in file-sharing. Computer Communication & Distributed Systems EDA 390 The Role and uses of Peer-to-Peer in file-sharing Computer Communication & Distributed Systems EDA 390 Jenny Bengtsson Prarthanaa Khokar jenben@dtek.chalmers.se prarthan@dtek.chalmers.se Gothenburg, May

More information

Load Balancing and Switch Scheduling

Load Balancing and Switch Scheduling EE384Y Project Final Report Load Balancing and Switch Scheduling Xiangheng Liu Department of Electrical Engineering Stanford University, Stanford CA 94305 Email: liuxh@systems.stanford.edu Abstract Load

More information

51-10-50 Circuit-Switched Router Connections Nathan J. Muller

51-10-50 Circuit-Switched Router Connections Nathan J. Muller Previous screen 51-10-50 Circuit-Switched Router Connections Nathan J. Muller Payoff LAN managers will find that routers supporting dial backup, bandwidth-on-demand, and dial-on-demand enable more flexible

More information

How Network Transparency Affects Application Acceleration Deployment

How Network Transparency Affects Application Acceleration Deployment How Network Transparency Affects Application Acceleration Deployment By John Bartlett and Peter Sevcik July 2007 Acceleration deployments should be simple. Vendors have worked hard to make the acceleration

More information

Effects of Filler Traffic In IP Networks. Adam Feldman April 5, 2001 Master s Project

Effects of Filler Traffic In IP Networks. Adam Feldman April 5, 2001 Master s Project Effects of Filler Traffic In IP Networks Adam Feldman April 5, 2001 Master s Project Abstract On the Internet, there is a well-documented requirement that much more bandwidth be available than is used

More information

P2P VoIP for Today s Premium Voice Service 1

P2P VoIP for Today s Premium Voice Service 1 1 P2P VoIP for Today s Premium Voice Service 1 Ayaskant Rath, Stevan Leiden, Yong Liu, Shivendra S. Panwar, Keith W. Ross ARath01@students.poly.edu, {YongLiu, Panwar, Ross}@poly.edu, Steve.Leiden@verizon.com

More information

Content Delivery Networks

Content Delivery Networks Content Delivery Networks Terena 2000 ftp://ftpeng.cisco.com/sgai/t2000cdn.pdf Silvano Gai Cisco Systems, USA Politecnico di Torino, IT sgai@cisco.com Terena 2000 1 Agenda What are Content Delivery Networks?

More information

One of the most important topics in any discussion of TCP/IP is IP. IP Addressing

One of the most important topics in any discussion of TCP/IP is IP. IP Addressing IP Addressing 125 machine, called a RARP server, responds with the answer, and the identity crisis is over. RARP uses the information it does know about the machine s MAC address to learn its IP address

More information

LTE BACKHAUL REQUIREMENTS: A REALITY CHECK

LTE BACKHAUL REQUIREMENTS: A REALITY CHECK By: Peter Croy, Sr. Network Architect, Aviat Networks INTRODUCTION LTE mobile broadband technology is now being launched across the world with more than 140 service providers committed to implement it

More information

July, 2006. Figure 1. Intuitive, user-friendly web-based (HTML) interface.

July, 2006. Figure 1. Intuitive, user-friendly web-based (HTML) interface. Smart Switches The Value-Oriented Alternative for Managed Switching White Paper September, 2005 Abstract This White Paper provides a short introduction to Web Smart switches and their importance in a local

More information

Compact Representations and Approximations for Compuation in Games

Compact Representations and Approximations for Compuation in Games Compact Representations and Approximations for Compuation in Games Kevin Swersky April 23, 2008 Abstract Compact representations have recently been developed as a way of both encoding the strategic interactions

More information

Application Performance Monitoring (APM) Technical Whitepaper

Application Performance Monitoring (APM) Technical Whitepaper Application Performance Monitoring (APM) Technical Whitepaper Table of Contents Introduction... 3 Detect Application Performance Issues Before Your Customer Does... 3 Challenge of IT Manager... 3 Best

More information

VOICE OVER IP AND NETWORK CONVERGENCE

VOICE OVER IP AND NETWORK CONVERGENCE POZNAN UNIVE RSITY OF TE CHNOLOGY ACADE MIC JOURNALS No 80 Electrical Engineering 2014 Assaid O. SHAROUN* VOICE OVER IP AND NETWORK CONVERGENCE As the IP network was primarily designed to carry data, it

More information

Content Delivery Networks (CDN) Dr. Yingwu Zhu

Content Delivery Networks (CDN) Dr. Yingwu Zhu Content Delivery Networks (CDN) Dr. Yingwu Zhu Web Cache Architecure Local ISP cache cdn Reverse Reverse Proxy Reverse Proxy Reverse Proxy Proxy L4 Switch Content Content Content Server Content Server

More information

QoS Issues for Multiplayer Gaming

QoS Issues for Multiplayer Gaming QoS Issues for Multiplayer Gaming By Alex Spurling 7/12/04 Introduction Multiplayer games are becoming large part of today s digital entertainment. As more game players gain access to high-speed internet

More information

Demand Routing in Network Layer for Load Balancing in Content Delivery Networks

Demand Routing in Network Layer for Load Balancing in Content Delivery Networks Demand Routing in Network Layer for Load Balancing in Content Delivery Networks # A SHRAVANI, 1 M.Tech, Computer Science Engineering E mail: sravaniathome@gmail.com # SYED ABDUL MOEED 2 Asst.Professor,

More information

IP-VPN Architecture and Implementation O. Satty Joshua 13 December 2001. Abstract

IP-VPN Architecture and Implementation O. Satty Joshua 13 December 2001. Abstract Abstract Virtual Private Networks (VPNs) are today becoming the most universal method for remote access. They enable Service Provider to take advantage of the power of the Internet by providing a private

More information

Global Server Load Balancing

Global Server Load Balancing White Paper Overview Many enterprises attempt to scale Web and network capacity by deploying additional servers and increased infrastructure at a single location, but centralized architectures are subject

More information