Lec 1 Cloud Introduction.pdf

Clouds 1 Introduction Fundamentals of Network and Cloud Computing Dr Dawn Carmichael Today  Definitions  Five essential characteristics  Virtualisation  Explanatory case studies  Cloud providers Background Definitions  A distributed computer system is a widely dispersed collection of computers...

Clouds 1 Introduction Fundamentals of Network and Cloud Computing Dr Dawn Carmichael Today  Definitions  Five essential characteristics  Virtualisation  Explanatory case studies  Cloud providers Background Definitions  A distributed computer system is a widely dispersed collection of computers that appears to the user as a single coherent system  Grid computing is a type of distributed system where the system acts as a super virtual computer  Cloud computing is a type of distributed system where the system provides services to a shared pool of resources Cloud computing definitions  A model of on demand network access to shared computing resources such as networks, servers, storage, applications, data, analytics and services  A colloquial term with many loose definitions  Overloaded with meanings National Institute of Standards and Technology (NIST) Definition “Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model promotes availability and is composed of five essential characteristics, three service models, and four deployment models.” Cloud computer from the consumer point of view  Accessible via a browser or a web service API  Don’t have to buy computer hardware/software  Pay for what you use Grid Computing vs Cloud Computing GRID CLOUD  Application oriented  Service oriented  Tasks are divided into sub-task and allocated to machines  Provides services on demand  Interconnected computers working on a large scale task  Usually owned by an organisation within a corporate network  Management is de-centralised  Accessed via internet  Owned by an infrastructure provider  Management is centralised Cloud Essential Characteristics  On-demand self service – sign up and service access  Broad network access – PC, laptops, mobile  Multi-tenancy / Resource pooling – across consumers  Rapid Elasticity – can scale and cope with demand spikes  Measured service – metered charging of service abstractions (storage, processing etc.) Virtualisation & Cloud Computing  The main enabling technology for Cloud Computing is Virtualization.  Virtualization is, for example, a partitioning of single physical server into multiple logical servers.  Once the physical server is divided, each logical server behaves like a physical server and can run an operating system and applications independently.  A hypervisor is software which runs virtual machines Multi-tenant Model Hypervisor enables this model Virtualisation  Separates resources and services from the physical system  Partitioning – single physical system providing many applications and OS’s  Isolation – each VM isolated from other VMs. If one crashes it doesn’t affect others  Encapsulation – VM can be presented to an application as an entity. If one application crashes it doesn’t affect others Virtualisation Types  Network Virtualization: It is a method of combining the available resources in a network by splitting up the available bandwidth into channels, which can be assigned to a specific server or device in real time.  Storage Virtualization: It is the pooling of physical storage from multiple network storage devices into what appears to be a single storage device for example a Storage Area Networks (SANs).  Server Virtualization: Server virtualization is the masking of server resources like processors, RAM, operating system etc. - to increase the resource sharing and reduce the burden and complexity of computation from users. Hypervisor  Software that runs VMs – basically an OS  Close to the hardware so that it can facilitate multiple operating systems  It schedules hardware resources for other OSs  Treats OS the same as applications  Allows the same application on different OS VM hypervisor VM hypervisor VM Hypervisors Types  Type 1 - Native – sit directly on the hardware – MS Hyper-V  Type 2 - Hosted – run as software on other hardware and OS – VMWare player Abstracting Hardware  File system virtualisation – VMs access different files systems  Virtual symmetric multi-processing – VMs can access multiple processors simultaneously  Virtual high-availability support – VM crashes auto restart on another server  Distributed Resource Scheduler – manages multiple hypervisors across hardware VM hypervisor VM DRS hypervisor VM Virtualisation Types  Static  Silo – isolated software and hardware – always the same VM  Single server or server group - relatively simple  Significantly more efficient that no virtualisation – scalability  Not optimal use of server or server group  Dynamic  Allocate servers dynamically – user profile mapped to a new VM  More complex  Optimal use of server or server group Desktops in the cloud  Client Virtualisation  Applications, data files etc. not stored on a PC stored on a server  Motivation – Total Cost of Ownership (TCO)  purchase and maintenance of PCs  out of date within five years typically  Virtual desktop may bring down TCO Client Virtualisation Types  Session based – server is running single OS with multiple sessions for users  Operating system streaming – OS is passed to users as required i.e. streaming – some processing on the server some on the client  Virtual Desktop Infrastructure – server runs virtual PCs sent to user – VMs on the server sent to the users  PC blade – a server blade is server on one computer board slotted into a blade cabinet – a server blade can contain a number of PCs which can be allocated to users Case study – email service GENERAL USERS Traditional on premises email  Run in the organisation  Requires  Server  Server software  Personnel  Issues  Will require estimates of present and future usage  Fault tolerance has to be provisioned  Will be redundant within 3-5 years Cloud E-mail  Email providers – Gmail, outlook, protonmail, zoho  Requires  Choose package – maybe ‘free’ folded in with web services  Doesn’t require capital outlay  Scalable  Fault tolerant  Issues  Outsources privacy and security Case study – file storage service UNIVERSITY STUDENTS Traditional on premises – virtual drive  Run in the organisation  Requires  Server  Server software  Personnel  Issues  Will require estimates of present and future usage  Fault tolerance has to be provisioned Cloud storage  Requires  Choose package – maybe ‘free’  Doesn’t require capital outlay  Scalable  Fault tolerant  Issues  Outsources privacy and security Case study – web site hosting BUSINESS Traditional on premises – Web site deployment  Developed and hosted in the organisation  Requires  Server  Server software  Personnel  Web development  Issues  Will require estimates of present and future usage  Fault tolerance has to be provisioned Cloud – web sites  Options  Web development platforms – Wix, GoDaddy, Wordpress  Web Site Hosting – Google Cloud Platform, MS Azure, Amazon Web Services (AWS)  Doesn’t require capital outlay  Scalable  Issues  Lack of strategy can be a problem – back-end on GoDaddy, Wordpress UI Cloud Providers Cloud Providers Key offerings for each – not an exhaustive list Cloud Providers Market Share Top Four Cloud Service Providers Paradigm Shift Summary