Plataformas de Suporte a IA

Objectives

Students should be able to: - Understand with a comprehensive approach the full spectrum of modern IT operations (conception, design, configuration, deployment, administration and management); - Understand the technological evolution of IT infrastructures, on-premises and cloud-based, with an emphasis on platforms for IA systems; - Master the principles underlying the virtualization of IT infrastructures; - Bridge software development techniques and the provisioning and deployment mechanisms of those IT infrastructures; - Know the business concepts, models, and strategies of IT services, and how to apply the best practices of administration and management of IT infrastructures; - Understand the role of IT service control and management frameworks in the management of large scale IT infrastructures; - Minimize the impact of IT infrastructures on the environment and on resource depletion.

Syllabus

Virtualization: trap and emulate, virtualized memory and I/O; virtualization solutions and their tradeoffs: virtual machines, containers, unikernels, microVMs; Hardware AI accelerator virtualization; Infrastructure-as-Code: programmatically provisioning infrastructure; declarative vs imperative infrastructure; idempotent and immutable infrastructure; Cloud computing: main characteristics; deployment models and service models (including Machine- Learning-as-a-Service); multi-cloud and edge computing; Deployment and Orchestration: difference between deployment and orchestration; different types of deployment; introduction and discussion of popular orchestrators such as Docker Swarm and Kubernetes; Cloud Native: microservice architectures; functional decomposition, data partitioning; CAP theorem; event sourcing; Function-as-a-Service: function programming model, serverless infrastructure management, dealing with cold starts; DevOps: CI/CD; Jenkins; GitOps; MLOps: CI/CD for ML applications

Prerequisites

Students must have programming knowledge and be able to use languages such as Python and Java.

Cross Competence Component

Laboratory experiments carried out and project developed by groups of 2 or 3 students (weight in grading:50%), thereby spurring several transversal skills: - group planning and management of activities - decomposition of complex tasks into smaller work units menores that can be carried out independently - development of mutual help and group-member accountability skills - discussion among the group members in order to reach consensus on the best solutions, improving the argumentative and communication abilities - knowledge sharing and perspective contrasting among the group members

Laboratorial Component

Laboratory experiments using automation scripts and tools to build and manage IT infrastructures, on-premises and in cloud-computing platforms.

Programming And Computing Component

The course where this UC is integrated ensures the components of Computation and Programming according to the MEPP 2122.

Ethical Principles

All members of a group are responsible for the group’s work In any assessment, every student shall honestly disclose any help received and sources used. In an oral assessment, every student shall be able to present and answer questions about the entire assignment and solution.