Cloud Computing Advanced and Emerging Cloud Topics Syllabus

Advanced and Emerging Cloud Topics syllabus — full chapter and topic list

Expand any chapter to see its topics and sub-topics. This is the whole examinable outline for Advanced and Emerging Cloud Topics in Cloud Computing, not a summary of it.

1. 1

   Big Data and Analytics in the Cloud

   4 topics

   • Data Lakes and Lakehouses
   • Managed Spark and Hadoop
   • Stream Processing
   • Business Intelligence Services
2. 2

   Machine Learning and AI Services

   4 topics

   • Managed ML Platforms
   • Pre-Trained AI APIs
   • Generative AI and Foundation Models
   • MLOps
3. 3

   Edge and IoT Computing

   4 topics

   • Edge Computing Fundamentals
   • IoT Device Management
   • Fog Computing
   • Latency-Sensitive Workloads
4. 4

   Multi-Cloud and Hybrid Operations

   4 topics

   • Cloud Abstraction Layers
   • Hybrid Connectivity (Anthos, Azure Arc)
   • Workload Portability
   • Avoiding Vendor Lock-In
5. 5

   Sustainability and Green Cloud

   3 topics

   • Carbon-Aware Computing
   • Energy-Efficient Architectures
   • Sustainability Reporting

Advanced and Emerging Cloud Topics flashcards for Cloud Computing

25 of 50 cards from the Advanced and Emerging Cloud Topics deck — real questions with worked answers.

1. What is a data lake, and how does its schema handling differ from a traditional data warehouse?

   A data lake is a centralized repository that stores vast amounts of raw data in its native format (structured, semi-structured, unstructured). It uses schema-on-read (schema applied when data is queried), whereas a data warehouse uses schema-on-write (schema enforced at ingestion).

2. What is a data lakehouse, and what two paradigms does it combine?

   A lakehouse is an architecture that combines the low-cost, flexible storage of a data lake with the ACID transactions, schema enforcement, and BI performance of a data warehouse, typically via a metadata/transaction layer over cheap object storage.

3. Name three open table formats that enable ACID transactions on data lakes (lakehouse foundations).

   Delta Lake, Apache Iceberg, and Apache Hudi.

4. In a lakehouse, what functions does the transactional metadata layer (e.g., Delta/Iceberg) provide over raw object storage?

   ACID transactions, schema enforcement and evolution, time travel (versioned snapshots), efficient upserts/deletes, and metadata indexing for query performance.

5. What is the difference between Apache Hadoop MapReduce and Apache Spark in processing model?

   MapReduce writes intermediate results to disk between each map/reduce stage; Spark performs in-memory processing using RDDs and a DAG execution engine, making it typically much faster for iterative and interactive workloads.

6. What are the two core components of the traditional Apache Hadoop framework?

   HDFS (Hadoop Distributed File System) for storage and MapReduce (with YARN as resource manager) for distributed processing.

7. Name the major managed Spark/Hadoop services on AWS, Azure, and Google Cloud.

   AWS EMR (Elastic MapReduce), Azure HDInsight (and Synapse/Databricks), and Google Cloud Dataproc. Databricks is a leading cross-cloud managed Spark platform.

8. In Apache Spark, what is an RDD and what key property makes it fault-tolerant?

   An RDD (Resilient Distributed Dataset) is an immutable, partitioned collection of records. Fault tolerance comes from lineage: Spark records the sequence of transformations so lost partitions can be recomputed.

9. What is the difference between batch processing and stream processing?

   Batch processing operates on large, bounded datasets collected over time and processed together; stream processing handles unbounded data continuously, record-by-record or in micro-batches, with low latency.

10. Distinguish event time from processing time in stream processing.

    Event time is when the event actually occurred (embedded timestamp); processing time is when the system observes/processes it. The gap between them is called skew or lag and drives the need for watermarks and windowing.

11. What is a watermark in stream processing?

    A watermark is a heuristic marker asserting that no events with an event time earlier than a given threshold are expected anymore, allowing the system to finalize windows and handle late-arriving data.

12. Compare tumbling, sliding, and session windows in stream processing.

    Tumbling windows are fixed-size, non-overlapping intervals; sliding windows are fixed-size but overlap and advance by a step smaller than the window; session windows are dynamic, grouping events separated by gaps shorter than a timeout.

13. Name common managed stream-processing services across the major clouds.

    AWS Kinesis Data Streams / Managed Service for Apache Flink; Azure Event Hubs / Stream Analytics; Google Cloud Pub/Sub / Dataflow (Apache Beam). Apache Kafka and Flink are the open-source foundations.

14. What delivery guarantees can streaming systems provide, from weakest to strongest?

    At-most-once (may drop, never duplicate), at-least-once (never drop, may duplicate), and exactly-once (each record affects state precisely once, typically via idempotency or transactional checkpoints).

15. What is a cloud Business Intelligence (BI) service and what is its primary purpose?

    A managed BI service provides data visualization, dashboards, and reporting over connected data sources, enabling self-service analytics and decision-making. Examples: Amazon QuickSight, Power BI, Google Looker/Looker Studio.

16. In BI/data-warehouse design, contrast a star schema with a snowflake schema.

    A star schema has a central fact table linked directly to denormalized dimension tables; a snowflake schema normalizes those dimensions into multiple related sub-tables, reducing redundancy but adding join complexity.

17. What is OLAP versus OLTP?

    OLAP (Online Analytical Processing) is optimized for complex, read-heavy analytical queries over historical data (BI); OLTP (Online Transaction Processing) is optimized for high-volume, short read/write transactions (operational apps).

18. What is a managed ML platform and what lifecycle stages does it typically cover?

    A managed ML platform (e.g., Amazon SageMaker, Azure Machine Learning, Google Vertex AI) provides tooling across the ML lifecycle: data labeling/prep, training, hyperparameter tuning, model registry, deployment/serving, and monitoring.

19. On the major clouds, name the flagship end-to-end managed ML platforms.

    AWS SageMaker, Microsoft Azure Machine Learning, and Google Cloud Vertex AI.

20. What are pre-trained AI APIs, and give three common categories.

    Pre-trained AI APIs are ready-to-use cloud services exposing trained models via API without needing to train them. Common categories: vision (image/object recognition), speech (speech-to-text/text-to-speech), and natural language (translation, entity/sentiment analysis).

21. Give examples of pre-trained vision, speech, and language APIs across the clouds.

    Vision: AWS Rekognition, Azure Computer Vision, Google Vision AI. Speech: AWS Transcribe/Polly, Azure Speech, Google Speech-to-Text. Language: AWS Comprehend/Translate, Azure Language, Google Natural Language/Translation.

22. What is a foundation model?

    A foundation model is a large model trained on broad, unlabeled data at scale that can be adapted (via fine-tuning or prompting) to a wide range of downstream tasks. LLMs like GPT and Claude are examples.

23. What is the difference between fine-tuning and prompt engineering (in-context learning)?

    Fine-tuning updates a model's weights by training on task-specific data; prompt engineering / in-context learning guides a frozen model's behavior purely through the input prompt and examples, without changing weights.

24. What is RAG (Retrieval-Augmented Generation) and why is it used with foundation models?

    RAG augments a generative model by retrieving relevant documents (often via vector similarity search) and adding them to the prompt as context. It grounds responses in up-to-date, domain-specific data and reduces hallucination without retraining.

25. Name the managed generative-AI / foundation-model services on AWS, Azure, and Google Cloud.

    AWS Bedrock, Azure OpenAI Service (and Azure AI Foundry), and Google Vertex AI (Gemini/Model Garden).

See more Advanced and Emerging Cloud Topics flashcards →

Planning Advanced and Emerging Cloud Topics for Cloud Computing

Advanced and Emerging Cloud Topics (Cloud Computing) FAQ