Parallel and Distributed Computing Chapter 6 Cloud Computing

Parallel and Distributed Computing Chapter 6 Cloud Computing

6.1 What is cloud computing

Cloud computing is the provision of computing services (including servers, storage, databases, networks, software, analysis and intelligence) through the Internet ("cloud"). You only need to purchase on-demand from the cloud service provider and pay as you go; you no longer need to pay attention to everything. The "cloud" hides all the details of the purchase, connection and maintenance of physical hardware for you. The cost on the cloud is lower than building it yourself.

Elastic Computing
Elastic computing quickly scales up or down a computer's processing, memory, and storage resources to meet changing demands, eliminating the need for capacity planning and engineering that has to worry about peak usage . This is a very attractive feature that cloud computing can provide.

Elastic computing is often controlled by system monitoring tools that match the amount of resources allocated to the amount of resources actually required without interrupting operations.

With elastic computing, you avoid paying for unused capacity or idle resources, and you don't have to worry about investing money to purchase or maintain additional resources and equipment.

Fault tolerance, high availability, disaster tolerance
• Fault tolerance is the property that enables a system to continue to operate normally despite the failure of one or more failures in some of its components. System performance may be degraded (business is not interrupted, generally used in life-critical systems)

• The purpose of high availability is to keep the system operating at a certain level for a longer period of time, that is, to reduce downtime as much as possible (allowing short business interruptions)

• Disaster recovery refers to the system's ability to restore business when a disaster occurs. Generally, disaster recovery is used to protect the most critical business data (data backup)

Cloud computing service model
IaaS (Infrastructure as a Service)
Infrastructure: all physical-level hardware resources, such as computing cores (CPU, GPU, FPGA), storage, network, etc.

• Elastic computing: Amazon EC2, Aliyun ECS…
• Block storage, object storage, etc.
• Private network

SaaS (Software as a Service)
Provide the application itself directly over the network
Generally accessible through a browser

• Cloud office (various cloud documents)
• ERP
• Cloud gaming

PaaS (Platform as a Service)
Between IaaS and SaaS. Based on IaaS, basic software such as OS, middleware, and DB are also provided as services directly to you. A complete runtime environment PaaS

6.2 Virtualization technology

You may have used a virtual machine. It is a kind of virtualization technology. The PC/Mac physical machine we usually use is generally composed of two layers: physical hardware + operating system. The virtual machine uses software to simulate hardware functions and runs on the host machine. Multiple virtual operating systems and applications

What virtualization can do

  • Partition
    • Can run multiple operating systems on one physical machine,
    • Can allocate system resources among virtual machines
  • Isolation
    • Enables fault and safety isolation at the hardware level
    • Maintains performance with advanced resource control capabilities
  • Encapsulation
    • Saves the complete state of a virtual machine to a file.
    • Moving and copying virtual machines is as easy as moving and copying files.
  • Hardware Independent
    • Any virtual machine can be provisioned or migrated to any physical server.

Guess you like

Origin blog.csdn.net/weixin_61197809/article/details/134623885