SOA-based cloud computing PAAs platform in the logistics field

1. The dilemma of enterprises in logistics field

In the competition of many logistics enterprises, whether the enterprise can stand out, depends on how quickly and efficiently adapt to the change of the market. A company that wants to be able to adapt quickly to change needs a flexible system that maximizes proximity to customers, responds to customer dynamic needs, and helps companies seize dynamic business opportunities. Therefore, the business processing of enterprises must go out of their own scope, with multiple customers and partners to coordinate. For example. The business of an international freight forwarding company, usually across the boundaries of the enterprise, its role in addition to internal operations, documents, customer service, finance, sales, managers, but also includes the entire supply chain upstream and downstream enterprises, such as shippers, peers, convoys, customs broker, overseas agents, consignee and so on. In traditional it mode, it is not possible to automate services in these business across the enterprise. Customers must manually send orders, check inventory, and send email or phone calls to suppliers, all of which are well implemented automatically under the Cloud computing platform. From the perspective of business people, cloud computing is not an Enterprise Portal system, not a supply chain management system, but a business circle and value-added chain, is a business with customers, enterprises and cooperative enterprise social network.

In the traditional logistics industry, people implement the logistics information management system. Logistics distribution system, invoicing and inventory management system. Most are built in the form of an internal system. At the same time, deploying these applications is a very complex, expensive and risky business. You need to assemble the hardware, the operating system, for each application. Database。 Middleware, Web servers, and other software. Each logistics enterprise needs to train a team of experts including network, database, System management to ensure the normal operation of the application software. Once the new business needs to change the original application system, a long development process begins again and again.

2. Cloud computing in the field of logistics applications

Logistics in the field of cloud computing, can let the logistics enterprises according to their actual size and demand, dynamically from the cloud of the Internet to choose the corresponding visualization of resources and services to meet the enterprise in the day-to-day operation of the various IT services needs.

IBM's intelligent "logistics cloud" presents a similar concept. It provides an intelligent logistics solution based on cloud computing technology, which can be used in the field of logistics, and will comprehensively improve the intelligence level of cargo handling, transportation, warehousing, overhaul and customs clearance, and realize efficient, fast, intensive and transparent logistics industry, and save management cost. Improve the management level. As a platform for cloud software services and applications development (Platform as a serice), it provides a platform for building and running software services while, on the other hand, manages all hardware and software resources, providing customers on demand via the Internet, web-based Software Solutions.

PAAs provides all the required running applications in the Internet infrastructure. Users simply "Turn on the tap" to get the service, they don't have to worry about the complexity behind the scenes. PAAs is based on subscription mode, so users only need to pay for the features he uses. With PAAs, independent software developers and enterprise IT departments can focus more on innovation than on complex infrastructure. Logistics companies can devote more of their budgets to places where they can provide real business value, rather than the purchase and maintenance of basic equipment.

3. Cloud computing system and Platform development status

At present, cloud computing infrastructure or cloud computing platform, open source organization and academia have put forward many cloud computing systems or platform solutions from Amazon, Google, IBM, Microsoft, Sun and other companies.

3.1 Google's cloud infrastructure

Google's cloud computing infrastructure is progressively expanded on the basis of initial service offerings for search applications, mainly Distributed file systems (Googlefile system (GFS), large scale distributed database BigTable, programming mode Mapreduee, The distributed locking mechanism is composed of several independent and tightly chubby systems.

3.2 IBM "Blue Cloud" computing platform

IBM's Blue Cloud computing platform is made up of a data center, IBM Tivoli Monitoring Software (Tivoli monitoring), IBMDB2 database, IBM Tivoli Deployment Management software (Tivoli provi-sioning Manager), IBM WebSphere Application Server, and open source virtualization software together with some open-source information processing software.

3.3 Sun's cloud infrastructure

The cloud infrastructure architecture proposed by Sun includes 6 levels of services, applications, middleware, operating systems, virtual servers, physical servers, and offers the idea that cloud computing can describe services delivered at any traditional level from hardware to applications.

3.4 Microsoft's Azure cloud Platform

Microsoft's Azure Cloud Platform consists of 4 tiers at the bottom of Microsoft's Global Foundation Service,gfs, consisting of a fourth generation of data centers across the globe: Cloud infrastructure services (cloud infrastructure Service, with Windows Azure operating system as the core, primarily engaged in virtualized computing resource management and Intelligent task assignment: On top of Windows Azure is an application service platform that plays the role of a component (building block), To provide users with a range of services.

3.5 Amazon's Elastic computing cloud

Amazon is one of the first companies to provide cloud computing services. The company's flexible computing Cloud (elastic compute CLOUD,EC2) platform is built on a large scale of computer and server clusters within the company, providing users with the various virtual machine instances (instance) that the network interface operates in "cloud" operation.

4. SOA-based cloud computing platform framework

Each cloud computing platform also has different characteristics, especially in the use of platform, transparent computing platform for users to provide users with the actual contact of the client nodes and remote virtual Storage server can not contact. is a semi public environment, Google's cloud computing platform environment is a private environment. In addition to opening limited application interfaces, such as GWT (Googlewebtoolkit), Googleappengine, and Googlemapapi, Google does not share the cloud's internal infrastructure with external users, and IBM's "Blue Cloud" The computing platform is a collection of hardware and software available for sale. Users build their own cloud computing applications based on these software and hardware products, and Amazon's flexible Computing cloud is a hosted cloud computing platform that allows users to operate directly using the remote interface. The actual physical node is not visible. From other points of view of the different cloud computing systems, it can be seen that, although the cloud computing system in many ways has a common, but in fact, the system is still very different from each other, which gives cloud users or developers a different experience.

For these cloud computing platforms, we are designing the architecture of the SOA based cloud computing platform. Will include multiple levels of hardware and system software. Generally, it can be divided into the following three levels:

4.1 Hardware Platform

Hardware platform is commonly known as IaaS, it is mainly to provide users with virtualized computer resources, storage resources, network resources. All hardware facilities, including servers, network devices, storage devices, and so on, it is the data center of cloud computing. The hardware platform must first have the extensibility fscaling, the user can assume that the hardware resources are infinitely many. According to their own needs, users dynamically use these resources, and according to the amount of use to pay service fees. You don't have to worry about how many devices you need to buy to support your current traffic.

In the design of hardware platform virtual technology is particularly important, it can allow multiple operating systems to share a large hardware facilities, so that the hardware platform providers can flexibly provide various types of cloud platform hardware requirements. Common, fee-based virtual technology (e.g., VMware). There are also free open-source technologies (e.g. Xen).

4.2 Cloud Platform

The cloud platform here is specifically PAAs, which provides service development tools and basic software (e.g. databases, distributed operating systems, etc.). To help developers of cloud services. In addition, it is the operating platform of cloud services. Therefore, the cloud platform needs to have Java runtime, WEB2.O Application runtime, all kinds of middleware and so on.

4.3 Cloud services

Cloud services refers to one or more software features that can be accessed using a standard interface on the Internet. It is somewhat similar to the previous "Software as a service SaaS". But unlike SaaS. Traditional "Software as a service" system requires the service provider to provide and manage the hardware platform and system platform, while cloud services on cloud computing platform do not need to provide hardware platform and cloud platform. Customers can access and manage their business data by accessing all kinds of services over the Internet anytime, anywhere. Do not need to go to the customer site to install and debug software, configure the server and other operations.

Many vendors have already provided some of these platforms, such as IBM's smart Business Storage Cloud and Amazon's EC2, mainly a cloud-computing hardware platform (Hardware as a service), and Google's creator engine is primarily a cloud platform , Salesforce is a provider of cloud services.

In summary, through virtualization. The cloud computing platform is extremely flexible enough to meet a variety of requirements. not limited by hardware. In implementing your own cloud computing hardware platform, the main need to consider the storage structure, not only to consider the storage capacity, but also the need to consider the disk data read and write speed. The speed of a single disk is likely to limit the access of the service program to the data, so in practice it is necessary to distribute the data over multiple disks and to improve speed by reading and writing to multiple disks simultaneously. In addition, how data is placed is also a very important issue. Googlefilesystem's clustered file systems and block-based storage area network (SAN) systems provide two viable storage technologies. The Open source Hadoop HDFS (hadooddistributed File System) implements similar Googlefilesystem functions, providing a reference for a solution that implements the hardware platform.