Edge Computing Week 3 NPTEL Assignment Answers 2025

NPTEL Edge Computing Week 3 NPTEL Assignment Answers 2024

1. Which of the following can be considered as container orchestration functions,

• Provisioning container runtime on nodes
• Scheduling containers to run on hosts
• Creating a new container image for an application
• Restarting or Relaunching containers automatically on failure

Answer: a, b, d

Explanation:
Container orchestration involves automating the deployment, management, scaling, and networking of containers. Provisioning runtimes, scheduling containers, and handling failures are key orchestration tasks. Creating container images is a development task, not orchestration.

---

2. The ___________ plane is responsible for managing the state of the cluster while the ______________plane contains the worker-nodes that run the containerized application workloads.

• data, worker
• control, worker
• worker, data
• data, control

Answer: b

Explanation:
The control plane manages the cluster state, including scheduling and responding to cluster events. The worker plane (nodes) is where actual application workloads run.

---

3. Which of the following components in Kubernetes control plane is responsible for storing cluster state in a persistent manner?

• Controller Manager
• etcd
• API Server
• kube-proxy

Answer: b

Explanation:
etcd is a distributed key-value store used by Kubernetes to persist the entire cluster state, such as configuration data and metadata.

---

4. A “Service” in Kubernetes is used for:

• Providing external cloud-based services to the pod
• Load-balancing between pod replicas
• Communicating between pods
• Providing persistent data-storage to pods

Answer: b, c

Explanation:
Kubernetes Services abstract a set of pods and provide stable IP addresses and DNS names. They enable communication between pods and balance traffic across pod replicas.

---

5. In the NetworkTopologyStrategy in Cassandra, which of the following component is responsible for rack-awareness:

• Coordinator
• Snitch
• tombstone
• Bloom filter

Answer: b

Explanation:
A snitch in Cassandra determines the topology (like rack and data center) for replica placement. It helps in achieving fault-tolerant, rack-aware replication.

---

6. Replication factor (RF) in Cassandra refers to:

• Number of copies of a partition to be maintained
• Number of copies of memtable to be maintained
• Number of copies of SSTable to be maintained
• Number of copies of a snitch to be maintained

Answer: a

Explanation:
The Replication Factor defines how many copies of a partition exist in the cluster, increasing data redundancy and availability.

---

7. memtables in Cassandra are used to,

• store incoming write in memory
• read data from persistent storage
• log all the key-mutations
• perform the gossip protocol

Answer: a

Explanation:
Memtables are in-memory data structures that temporarily store incoming writes before they are flushed to disk as SSTables.

---

8. Cassandra provides strong Consistency. This statement is,

• True
• False

Answer: False

Explanation:
Cassandra prioritizes availability and partition tolerance (AP from CAP theorem), offering eventual consistency, not strong consistency by default.

---

9. Which of the following protocol is not usually used with IoTs?

• SMDP
• MQTT
• AMQP
• HTTPS

Answer: a

Explanation:
SMDP is not a standard IoT communication protocol. MQTT, AMQP, and HTTPS are commonly used in IoT for lightweight and secure communication.

---

10. Model Deployment phase in a ML workflow refers to,

• Collect data from sensors and devices
• Building and training AI/ML models
• Tracking model usage, analyse resources utilization and model related data
• None of the above

Answer: b

Explanation:
Model Deployment involves making the trained ML model available for predictions on new data — essentially integrating it into a production environment.