Strategies for Netflix's Distribution

Question 1: [10]

Netflix: Netflix uses three third party CDN providers to distribute their content namely – Akamai, Limelight and Level 3. The content sent to the client through adaptive streaming over HTTP. Netflix runs its service using virtual machines on Amazon Cloud. The functions within the Amazon Cloud include:

• Content Ingestion: the when the media is ingested and processed before being distributed to customers.
• Content processing: many formats of the media are created so that it suits the customers video players, whether its running on their phones, desktop or game consoles. This allows for adaptive streaming over HTTP using DASH.
• Uploading version to the CDNs: once all the formats are created, their uploaded to the designated CDNs

YouTube: This a video sharing site, that uses their own private CDN technology to distribute videos. With their servers done in a cluster, Google (Owner) uses DNS to redirect customers request to specific cluster where the RTT is lowest. YouTube uses HTTP streaming at the same time creating different formats for a video with different bit rate and quality. They also use HTTP byte range request to lessen the flow of data sent to the customer after a target amount of video had being prefetched.

KanKan: the media provider uses P2P instead of CDN delivery. P2P is quite similar to BitTorrent, whereby if a peer to download/view a video, it looks for other peers that may have a copy of that video. But the video is sent in chunks concurrently from other peers. The P2P uses a tracker and DHT technology to function.

Question 2: [9]

Streaming Live Audio/Video

• Allows user to receive a live radio or television transmission
• Many users receive the same video/audio at the same time
• Delay can be an issue

Streaming Stored Video

• Pre-recorded video are placed in servers
• User have control of how to view the video – play, pause etc.
• Uses UDP, HTTP and adaptive HTTP streaming
• Small Delay
• Uses Client-side Application buffer

The streaming technologies used in streaming stored video are – UDP streaming, HTTP streaming, adaptive HTTP streaming and Client-side buffering.

Question 3: [8]

Buffering helps to absorb the difference between server to client delay. For example, if video data is delayed and it being received by the client but not yet played, it will not be noticed even when the video itself is exhausted. Another reason why buffering is important is because if the server to client bandwidth falls below the video consumption rate, a user can view video playback as long as the buffer is not fully drained. The buffering that should be applied in video streaming service should be Client-Side Buffering.

Question 4: [7]

Question 5: [6]

The tasks involved in network management include:

• Project Management: this involves quantifying, measuring, reporting, analyzing and control of the performance of different network components.  Protocols such as SNMP (Simple Network Management Protocol) play an important role in project management.
• Fault Management: the main function of this type of management is to log detect and respond to fault conditions in the network. This involves quick handling of transient network failures like router hardware and software outages. The SNMP protocol also plays an important role in fault management.
• Configuration Management: this allows a network administrator/manager to track devices on the managed network including the hardware and software configuration on these network devices.
• Accounting Management: this enables the network administrator/manager to specify, log and control user and device access to network resources. This also involves usage quotas, allocation of resources access privileges for these devices too.
• Security Management: this controls access to network resources based on well-defined policy. This may involve the use of firewalls to monitor and control external access points to one’s network.

Question 6: [5]

The Internet-Standard Management Framework is a evolution of the SGMP (Simple Gateway Monitoring Protocol) which in turn allowed users to design implement and deploy SNMP within a few month.

The questions this is intended to address are:

• What is being monitored? And what kind of control can be exercised by the network administrator
• What is the specific form of the information that will be reported and/or exchanged?
• What is the communication protocol for exchanging this information?

The implication that the number of packets arriving at the leaky bucket is less that rt + b indicates that the number of packets moving within the network is below the maximum number of packets that can move within the network within any given time interval. This can also implicate that there are enough tokens are every packet moving within the network.

If the number of packets arriving at the leaky bucket is equal or greater than rt +b. Hence, then there will not be enough token to collected by the packets. Eventually the bucket itself will become empty and the packets will have to wait until the bucket has generated tokens to be collected by the packet. This essentially slows the flow of packets within the network.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more about our Essay Writing Service

In the video medium, each peer’s video stream is routed to a server and then from the server, the video stream is relayed back to each peer. Each peer receives N-1 streams from the N-1 other peers. Hence the number of bits per second the call initiator will need to send at rate r is (N-1) x (N-1) x r. The number bits N-1 callers will need to send is dependant on rate r bps.

Let us calculate the amount of time each class will get within the sequence

Class 1 Serving Time = 0.25/(0.25 + 0.5 + 0.75) = 0.25/1.5 = 16.7%

Class 2 Serving Time = 0.5/1.5 = 33.3%

Class 3 Serving Time = 0.75/1.5 = 50%

In order to get a sequence, we can use the lowest class weight (0.25) as a time unit, since its divisible to all other class weights. So, for example, class 2 has two serving times (0.5/0.25) with the sequence and class 3 has 3 serving times (0.75/0.25) within the sequence.

Hence the sequence can be (Class 1 = C1, Class 2 = C2, Class 3 = C3)  C1, C2, C2, C3, C3, C3…

Another sequence can be (In the sense of RR)  C1, C2, C3, C2, C3, C3 ….

Request Response Mode:

• Message is sent by management agent to managing entity
• Has more overhead
• Each piece of information requires two messages: poll and response

Trapping Mode:

• Message is sent by the managing entity, response coming from management agent
• Generates a single message to sender
• Polling has more overhead
• Only generates message when events occurs

Question 2.5: [1]

First, we convert the weight 178 to binary (2’s complement) which is  10110010, this is byte value 178.  The first byte transmitted stream has a value of 4 indicating data item to be OCTET string. The second byte will be the length of the OCTET string which is 7. The third byte contains the ASCII representation of the letter P. the T, L and V values of the next data item are 2 which the INTEGER tag value and 1 which the length of the integer (1 byte). Hence, the ENCODING would be  47Philips21178

Or it can transmitted by the byte stream as:

References

[1] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 9 pg. 781-782). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[2] http://subjects.ee.unsw.edu.au/tele4354/downloads/revCh9.pdf

[3] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 644). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[4] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 622-623). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[5] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 9 pg. 764-765). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[6] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 9 pg. 758-759). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[7] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 610-612). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[8] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 594). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[9] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 593-594). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248

[10] Kurose, J. F., Ross, K. W.  (07/2013, Ch. 7 pg. 608-612). Computer Networking: A Top-Down Approach, 6th Edition [VitalSource Bookshelf version].  Retrieved from vbk://978126939248