CS695 Topics in Virtualization and Cloud Computing
Spring 2019

Lecture 1  4.1.2019
-------------------

Introduction to virtualization and cloud computing

1. What is cloud computing and why the hype?

- on-demand scaling of services and infrastructure
- pay-as-you-use service model
- (massively) scalable over-the-network services
- internet-based computing

- computing-as-a-service
- analogies?

- computing requirements: 
  hardware, different abstraction/granulaties of 
  software services 

- main characteristics: on-demand, flexible/elastic provisioning, 
  fine-grained metering, networked.

- is this (service) model new?
  - conceptually, NO!
  - large centralized servers, cluster computing, grid computing, 
    utility computing etc., all provided some form of compute
    services as a service
  - now, due to
    decreasing costs and increased availability of powerful 
    feature-rich hardware and
    market and business models, the service based cloud computing
    model is back!

2. Typical service models
-     
   IaaS - infrastructure as a service
        - amazon ec2, hosting services offering VMs, dropbox, GPUclouds
   PaaS - platform as a service
        - appengine, container based hosting, VM appliances
   SaaS - software as a service
        - gmail, 
   FaaS - function as a service

   *aaS - acceleration as a service

- examples?

- private, public and hybrid cloud setups

- the different service models provide different levels of configuration
  and control. e.g., a machine vs. an email service.


3. Two-views of cloud services

User-view of cloud services
  - on-demand and elastic
  - quick deployment
  - secure and insolated environments
  - SLAs: performance, availability, reliability etc.
  - pay-on-use

Provider-view of cloud services
  - meet client SLAs
  - maximize resource sharing
  - reduce operation costs
  - demand forecasting and resource planning
  - dynamic resource allocation and costing

4. Problem areas and themes
- big data management and analytics
- programming models and tools
- security and privacy
- integration and interoperability
- regulation, compliance and economics
- energy management in data centers
- design and modeling of cloud-based applications

- mechanisms for service based models
  virtualization, containers, nested virtualization, lambdas,
  NFV, SDN, ...
- resource management mechanisms and policies in hosting environments
  memory, network, IO, ...
- enabling new capabilities with service models
  snapshoting, migration, high-availability, failovers and upgradets etc.

5. Did you mention virtualization in title?

  Yes! the scope of this course will mainly focus on issues related 
  to the infrastructure-as-a-service (IaaS) model.
  --- virtualization and containers as mechanisms for compute infrastructure.

6. Coming up,
   
   Exercise #1: qemu, kvm setup. VM installation. Linux source compile. 
   (due by 15th Jan).

   Exercise #2: Linux kernel module and system calls.
   (due by 30th Jan).

   Exercise #3: An extension/change/in-depth logging using the kvm hypervisor.
   (due by midsem).