NPTEL Cloud Computing Week 3 Assignment Answers 2024 (July-October)

This article provides detailed answers and explanations for the NPTEL Cloud Computing course's week 3 assignment for the July-October 2024 session. The assignment focuses on key concepts of cloud computing such as OpenStack storage management, parallel processing efficiency, SLA monitoring biases, the functionality of the Google File System, and the intricacies of service availability calculations. Additionally, the assignment delves into ephemeral storage, economies of scale in mobile services, resource provisioning in the face of exponential demand, common SLA parameters for different cloud services, and policies around data retention and deletion. Understanding these answers not only helps in acing the assignment but also deepens the learner's comprehension of cloud computing frameworks and their operational dynamics in real-world scenarios.

Answers and Explanations for NPTEL Cloud Computing Week 3 Assignment 2024 (July-October)

1. Which of the following statement(s) regarding OpenStack storage is/are right?

• Answer: A, C, D
  • A (Object storage is managed by Cinder): This is incorrect. Object storage is managed by Swift in OpenStack, not Cinder.
  • C (Block storage persists until VM is terminated): Incorrect, block storage can persist even after the VM is terminated unless specifically deleted. This allows reuse of storage across instances.
  • D (Ephemeral storage is used to run operating system and/or scratch space): Correct, ephemeral storage is typically used for the operating system and temporary storage that is directly attached to a running instance.

2. A task takes Time T in a uniprocessor system. In a parallel implementation, the task runs on P processors parallelly. The parallel efficiency is Eeff. What is the time taken by each processor (Tn) in this implementation?

• Answer: D. Tn = (T*Eeff) / P
  • This formula calculates the time taken by each processor considering the efficiency and parallelization. The task time T is divided by the number of processors P and adjusted for parallel efficiency Eeff.

3. What does the term "Bias towards vendors' empty" in the context of SLA monitoring?

• Answer: A. Vendor-driven selection of monitoring parameters
  • Bias towards vendors' empty indicates a tendency to select monitoring parameters that are more favorable or convenient for the vendor rather than being strictly aligned with customer needs or a balanced perspective.

4. How does the interface node in the Google File System maintain communication with chunk servers?

• Answer: D. Heartbeat messages
  • In Google File System (GFS), the master maintains communication and checks the status of chunk servers using periodic heartbeat messages.

5. In a cloud, total service uptime is 175 minutes and availability of the service is 0.85. What is the service downtime?

• Answer: C. 26.25 minutes
  • Downtime can be calculated using the formula: Downtime = Total time x (1 - Availability) = 175 x (1 - 0.85) = 26.25 minutes.

6. Regarding ephemeral storage:

• Answer: B. Statement 1 is TRUE, Statement 2 is FALSE
  • Ephemeral storage persists only until the VM is terminated (true), and it is not managed by Cinder in OpenStack (false; it's typically part of the local instance storage managed differently).

7. "Mobile service plans can achieve similar statistical economies to an infinitely large provider." Does this fall under?

• Answer: A. Correlated demand
  • Mobile service plans demonstrating correlated demand implies they can benefit similarly from economies of scale as an infinitely large provider, under certain statistical conditions.

8. Do (D) and (R) in the instantaneous demand and resources at time t respectively. If demand is exponential (D(t)=e^y), any fixed provisioning level (R) according to the current demands will fall linearly behind.

• Answer: A. TRUE
  • As demand grows exponentially and resources grow linearly or are fixed, the resource provision will increasingly fall short of meeting the demand.

9. Which of the following is/are expected common SLA parameters for both Software-as-a-Service and Storage-as-a-Service models?

• Answer: B. scalability
  • Scalability is a common SLA parameter for both SaaS and STaaS, reflecting the service's ability to scale resources up or down based on user demand.

10. Data retention and deletion by cloud providers do not fall under one of the SLA requirements.

• Answer: A. True
  • Generally, data retention and deletion policies are specified in data management policies or privacy policies rather than in SLAs, which typically focus more on performance and availability metrics.