Cloud Computing: Introduction and Overview

Overview

  • Instructor: Frédéric Fondement

  • Institution: Université Haute-Alsace

Page 1

  • Cloud computing refers to the delivery of computing services over the Internet, allowing for on-demand access to a shared pool of configurable resources such as servers, storage, applications, and services. This model promotes flexibility and cost-effectiveness.

Page 2: Reasons for Scalability

  • Big Data

    • Storage: The ability to efficiently store and retrieve vast volumes of data is essential as businesses generate enormous amounts of data daily. Efficient storage solutions help manage this influx, ensuring that data is readily available for use.

    • Analysis: Techniques for the rapid processing and analysis of large datasets are critical as organizations seek actionable insights from their data to drive business strategy and decision-making. Technologies such as data mining and machine learning play crucial roles here.

  • Web Applications/Services/Microservices

    • Numerous applications and services require the capacity to scale dynamically based on demand.

    • As user bases grow, the systems must maintain high levels of performance, ensuring reactivity, availability, and user satisfaction.

Page 3: Scalability Issue

  • How to React: Understanding and identifying appropriate response strategies is vital in addressing scalability issues effectively. This may involve short-term adaptations, such as increasing resources, or long-term strategic planning, including architectural changes.

Page 4: Tackling Scalability Issues

  • Profiling: A thorough analysis of resource utilization is necessary to identify bottlenecks and optimize performance. Key metrics to monitor include CPU, Memory usage, Disk IO, and Network IO. Profiling data helps organizations balance workloads and anticipate resource needs.

  • Monitoring: Ongoing monitoring is crucial for maintaining resource availability and assessing system performance against predefined service level agreements (SLAs). Automated monitoring tools can provide real-time insights into system health.

  • Caching: Implementing caching strategies, such as in-memory data stores, helps alleviate pressure on databases by storing frequently accessed data, thereby drastically improving response times and reducing load.

  • Data Structure Optimization:

    • Denormalization: In situations where performance is critical, applying NoSQL database concepts can offer advantages by allowing for more flexible data modeling tailored to current demands, facilitating faster queries and data retrieval speeds.

  • Improved Platforms: Leveraging advanced operating systems, programming languages, and database management systems allows for enhanced performance and scalability. It is essential to evaluate and adopt new technologies that align with business requirements.

  • Hardware: As scalability needs grow, investing in better physical infrastructure (such as enhanced servers) may be necessary to accommodate increased workloads and ensure optimized resource allocation.

Page 5: Scaling Solutions

  • Vertical Scalability: This approach involves enhancing the capacity of a single server or resource to handle increased load. For example, the National Supercomputing Center de Wuxi exemplifies this method with its significant computing capacity:

    • 40,960 processors, 10,649,600 cores, and an impressive performance of 93,014.6 Tflop/s, demonstrating effective vertical scaling.

Page 6: Horizontal Scalability

  • Horizontal scalability focuses on adding additional servers to distribute the load more evenly rather than solely upgrading existing servers. This approach enhances system resilience by reducing dependency on a single point of failure.

Page 7: Intellectual Framework of Cloud Computing

  • A comprehensive discussion on balancing resource allocation and elasticity across various physical setups is necessary for efficient cloud management. The XaaS (Everything as a Service) model represents a shift towards holistic service provisioning.

Page 8: Cloud Computing Concept

  • The concept of cloud computing encompasses the sharing of computation power and storage resources over a network, enabling scalable solutions that adapt to fluctuating demand.

Page 9-10: Web Applications in the Cloud

  • Classical Web Application: This term describes the typical deployment of web applications using cloud infrastructure, highlighting scalability and agility features.

  • Horizontal Scaling: Strategies for distributing load across multiple servers utilize protocols like HTTP/S to enhance system robustness and performance during peak times.

Page 11-12: Scaling Through Load Management

  • Stateless Processes: These processes are beneficial for scalable applications operating in multi-server environments since they don't retain session information, allowing for easy distribution of user requests across servers.

  • Addressing challenges related to shared caching, databases, and configurations must be approached with smart management strategies to avoid collaboration issues across different services.

Page 13: Resource Utilization

  • Connecting Resources: Aligning actual resource allocation with operational needs is essential for optimizing system performance. Automated adjustments and real-time monitoring are crucial for adapting to changing demands.

Page 14-15: Challenges in Scalability

  • Addressing linear scalability challenges involves tackling issues like coordination among servers, locking mechanisms for data integrity, and data sharing protocols during system failures.

  • Failures: The frequency of failures tends to rise with the addition of more hardware components, necessitating strategies for careful management and recovery processes to minimize downtime and service impact.

Page 16-19: Handling Failures and Monitoring

  • Monitoring Techniques: Effective resource usage tracking and chaos engineering practices can enhance understanding of system reliability and resilience.

  • Automating Processes: The shift towards automated management of processes aims to minimize human intervention and errors, leading to a more robust infrastructure management framework.

Page 20-24: Deployable Applications

  • There are specific requirements for applications to be considered horizontally and linearly scalable, fault-tolerant, and native to the cloud environment.

  • Tools: An introduction to existing platforms which include serverless computing and service orchestration capabilities to help meet these requirements while optimizing performance and resource management.

Page 25-26: Hardware and Network Structures

  • A thorough description of the physical framework of clouds includes components such as racks, servers, and their interrelationships within data centers. Understanding these relationships is key to effectively scaling and managing resources.

  • Supporting literature emphasizes the significance of both machine capabilities and network architectures in the successful deployment of cloud services.

Page 27-33: Data Center Revolution

  • Data Centers: Important specifications include aspects related to energy consumption, effective cooling solutions, and security measures pertinent to cloud frameworks. Understanding these factors supports the optimization of cloud scalability and operational efficiency.

  • The industry faces challenges related to carbon emissions due to increased energy consumption, necessitating the development of greener technologies.

Page 34-36: Cloud Computing Challenges

  • Key discussions focus on trends affecting internet traffic, workloads on cloud infrastructures, and metrics for energy usage and efficiency in data centers.

  • Rebound Effect: This term refers to the phenomenon where improved efficiency results in increased usage, further complicating resource management and sustainability efforts.

Page 37: Understanding Cloud Landscapes

  • A detailed comparison of private versus public cloud deployment models highlights their respective advantages and disadvantages, addressing key areas such as security, compliance, scalability, and cost.

  • This section provides a clarification on the varying service levels available in cloud computing, enabling better-informed decision-making by organizations.

Page 38-40: Cloud Services and Their Types

  • An overview of the cloud service models, including IaaS (Infrastructure as a Service), PaaS (Platform as a Service), and SaaS (Software as a Service) frameworks, providing insights into how these models function and their respective use cases.

Page 41-44: Transitioning from Private to Public Cloud

  • A detailed examination of the evolution from IaaS to PaaS illustrates real-world examples of cloud migration strategies, showcasing transitional challenges and solutions.

  • Discussion on how technologies like Kubernetes and serverless architectures facilitate efficient cloud management and reduce operational overhead.

Page 45-49: Evolution of Cloud Services

  • Insight into emerging cloud service models such as Function as a Service (FaaS) and Software as a Service (SaaS) underscores the shifting paradigms in software development and deployment methodologies.

Page 50-57: Conclusion on Cloud Computing

  • A comprehensive summarization of contemporary trends and practical experiences observed within cloud computing, highlighting its profound impact on the tech industry’s infrastructure.

  • Critical discussions surrounding operational elasticity, resource sharing, management practices, and future expectations in cloud-native applications pave the way for advancements and innovation.