UCS 531 Cloud Computing Exam Overview
Uploaded by
Panshul SaxenaUCS 531 Cloud Computing Exam Overview
Uploaded by
Panshul SaxenaCommon questions
Powered by AIStorage virtualization abstracts physical storage into a single storage pool, making management easier and more efficient, while network virtualization abstracts network resources and services, allowing for network optimization and flexibility .
IAM in Amazon secures EC2 by managing user access through key-pair authentication and configuring Security Groups for network permissions. For S3, IAM controls bucket and key permissions, ensuring proper access management and data protection within the cloud environment .
Google offers services such as Google App Engine, designed for scalable web services, while Amazon provides highly customizable options like EC2, S3, and Lambda for various computing needs. Services like complex ERP systems might be less suitable due to latency concerns, whereas consumer-facing applications benefit from the scalability of these platforms .
No-SQL databases, like MongoDB, support flexible schema and are optimized for unstructured data, providing better scalability and speed for specific workloads. Unlike SQL's structured query language and table-based structure, No-SQL employs collections of documents or key-value pairs, making it suitable for big data and real-time applications .
A Virtual Organization (VU) comprising companies A, B, and C could leverage shared resources and capabilities, such as digital modeling from B and spatial data from C, to enhance-market reach and operational efficiency. The graphical diagram would show interlinked shared attributes such as market insights and client databases, facilitating collaborative projects .
Google App Engine architecture supports automatic scaling and load balancing, enabling developers to deploy applications without managing underlying infrastructure. It integrates with services like Google Cloud Datastore and Google BigQuery to facilitate robust application development, leveraging Google's backbone for reliability and performance .
MapReduce algorithms facilitate processing large datasets by distributing tasks across nodes. Mappers process the input data, breaking it down into smaller fragments and outputting key-value pairs, while Reducers aggregate results to compute summary statistics like highest and lowest values per annum for large datasets, achieving scalability and efficiency .
In Hadoop, the Mapper function processes input data, emitting key-value pairs. The Combiner acts as a mini-reducer, aggregating intermediate data to reduce processing load, while the Reducer consolidates output from the Combiner, finalizing the data processing pipeline and ensuring comprehensive data analysis .
The HDFS ecosystem includes essential components like NameNode, which manages metadata and directory structure, and DataNodes, which handle actual data storage. Secondary NameNode assists in checkpointing to aid in recovery, while HDFS ensures data redundancy and fault tolerance through block replication .
Flynn's classification divides parallel computing into categories based on instruction and data streams: SISD (Single Instruction, Single Data), SIMD (Single Instruction, Multiple Data), MISD (Multiple Instruction, Single Data), and MIMD (Multiple Instruction, Multiple Data), each representing different computational scenarios and hardware design .