The IaaS service model of cloud computing architecture

IaaS

By IaaS This mode, users can get the computing or storage resources he needs from the vendor to load the relevant applications, and only pay for the portion of the resources they lease, and these cumbersome management tasks are assigned to the IaaS vendor.

1. History

Like SaaS, the idea of IaaS has been there for a long time, such as the past IDC (Internet Data Center, Internet Datacenter) and VPS (virtual private server, VPN), but because of technology, performance, The lack of price and use, these services are not widely used by large and medium-sized enterprises. But at the end of 2006, Amazon released the IaaS cloud service, EC2 (elastic Compute Cloud, a flexible computing cloud). Due to the advantages of EC2 technology and performance, such technology has been widely recognized and accepted by the industry, including some large enterprises, such as the famous New York Times.

2. Related Products

The most representative IaaS products are: Amazon EC2, IBM Blue Cloud, Cisco UCS and Joyent.

Amazon EC2. EC2 mainly to provide different specifications of computing resources (that is, virtual machines) mainly. It is based on the well-known Open-source virtualization technology Xen. With Amazon's various optimizations and innovations, EC2 has met enterprise-level requirements, both in performance and stability. It also provides a complete API and Web management interface for users to use.

IBM Blue Cloud. The Blue Cloud solution is the first in the industry developed by IBM's cloud Computing Center and is also a technologically advanced, enterprise-class cloud computing solution. The solution can consolidate the enterprise's existing infrastructure through virtualization technology and automated management technology to build their own cloud computing center, and to achieve the enterprise hardware resources and software resources Unified management, unified distribution, unified deployment, unified monitoring and unified backup, also broke the application of the monopoly of resources, So as to help enterprises to enjoy the cloud computing has brought many advantages.

Cisco UCS. It is the next-generation data center platform that integrates computing, networking, storage, and virtualization functions in a tightly coupled system. The system includes a low latency, no packet loss and support Gigabit Ethernet Unified Network array and multiple enterprise-class x86 architecture blade server and other devices, and in a unified management domain to manage resources. Users can support up to thousands of virtual machines by installing VMware vsphere on UCS. Cisco UCS enables enterprises to quickly build a cloud environment based on virtualization technology in local data centers.

Joyent. It provides IAAS services based on open Solaris technology. The core of its IaaS service is Joyent Accelerator, which delivers standards-based, non-proprietary, on-demand virtualized computing and storage solutions for Web application developers. Based on Joyent Accelerator, users can build their own network services with multi-core CPUs, massive memory and storage, and provide super fast access, processing speed and ultra-high reliability.

3. Advantages

IaaS services have a number of advantages over traditional enterprise data centers, the 5 most obvious below.

Maintenance-free. The main maintenance is handled by the IaaS cloud provider, so users don't have to worry.

Very economical. First of all, eliminates the user upfront hardware acquisition cost, and because the IaaS cloud mostly uses the virtualization technology, therefore the application and the server integration rate generally in 10 (namely one server runs 10 applications) above, this can reduce the usage cost effectively.

Open standards. Although many IaaS platforms have some proprietary functionality, IaaS is steadily progressing across platforms due to the advent of application publishing protocols such as OVF, which can be flexibly migrated across multiple IaaS clouds without being pinned to an enterprise data center.

Supported applications. Because IaaS mainly provides virtual machines, and the common virtual functions support a variety of operating systems, IaaS supports a wide range of applications.

Strong scalability. The IaaS cloud can provide users with a new computing resource in just a few minutes, while traditional enterprise data centers often take weeks, and computing resources can resize their resources based on user needs.

4. Technical

IaaS technology is a number of relatively low-level, of which 4 kinds of technology is more commonly used.

Virtualization. It can also be understood as a "multi-tenant" of the infrastructure layer. Because virtualization technology enables the generation of multiple virtual machines on a single physical server, and enables a complete isolation between these virtual machines, not only can the server's acquisition costs be reduced, but the server's operational costs are reduced. Mature x86 Virtualization technology has VMware ESX and open source Xen.

Distributed storage. In order to carry a large amount of data, but also to ensure that the data manageability, so a set of distributed storage System. In this regard, Google's GFS is a model.

relational databases. Basically, the original relational database based on the expansion and management of the optimization, so that it is more adaptable in the cloud.

NoSQL. To meet the goals that some relational databases cannot meet, such as supporting massive amounts of data, some companies have specifically designed databases that are not based on relational models, such as Google's BigTable and Facebook's Cassandra.

Most IaaS services are now based on Xen, such as Amazon's EC2, but VMware has also introduced vcloud based on ESX technology, while the industry has several cloud services based on relational databases, such as Amazon's RDS (relational Database service, relational databases service, and Windows Azure SDS (SQL data services,sql).

With regard to distributed storage and NoSQL, they have been widely used in the backend of the cloud platform, such as the datastore of Google App engine based on bigtable and GFS, while Amazon's simple DB is based on NoSQL technology.