Cloud computing environment to build a comprehensive and comprehensive automation framework

So whether you're an environmental deployment engineer or an automated scripting developer, reading this article will help you if you want to develop a solution for an automated environment deployment. The author of the project is IBM based on a large and complex cloud computing strategy products, by providing users with a visual network platform interface, fully and quickly to help customers deploy a high degree of flexibility of the cloud solution. In the process of project development and on-line, the average release of a new installation version every four hours, from software testing to production release requires more than four environments, especially in the demo (demo) environment every day, each involving redundant mirroring services, distributed architectures, encrypted VPN terminal access. While the deployment process is faced across geographies and across time zones, deployment packages are posted in the US Austin, Cloud Environment Data center in Singapore, where the deployment of personnel in Shanghai, China. The deployment process includes deployment of version packages for download and management, database upgrades, dynamic server program deployments, access to portal deployments, and static access to service component deployment and registration. If the traditional manual deployment, from beginning to end even skilled deployment engineers will need at least two days to complete, and manual deployment will occur omissions and errors occur, often encountered in the module leakage. So how to build a comprehensive and comprehensive automation framework, through the script to run the way to efficiently and accurately complete the deployment of the environment at all levels of the project is an important work progress. Now that the automation framework is online, it takes only one hour a day to deploy the latest version to all levels of the cloud computing environment accurately and efficiently. This automation framework has not only been applied to the project itself, its design ideas, functional modules are also applicable to other distributed, architecturally complex projects in the field of software engineering, so whether you are an environmental deployment engineer or an automated scripting developer, if you want to develop a solution for automated environment deployment, Reading this article will help you with your work.

This integrated solution includes at least the following benefits and features:

improve deployment efficiency of cloud computing environment eliminate human-made deployment omissions and errors, easily expand development tools open source, save costs

Design Architecture

First, we will start with the IBM Cloud Computing environment architecture to analyze the structure of the environment, so that we can specifically design a dedicated automation framework, the IBM cloud computing environment is based on IaaS (infrastructure as a service infrastructure services) to provide the cloud services, users through Web service obtains virtual facilities and bundled software services in the IBM cloud environment, such as users can apply for Linux, Windows and other operating systems on this platform, and can also apply for IBM Websphere, DB2 and other software products on this platform, which are stored in mirrored mode IBM Cloud environment.

To facilitate the reader's understanding of the context of the design and the interconnectedness of the internal modules, let's analyze the logical architecture diagram of the IBM cloud computing environment, as shown in the following illustration:

Figure 1. IBM Cloud computing Environment Logical architecture diagram

The user accesses the BSS (Business Service) platform of the IBM cloud environment through a WEB terminal, sending request-related mirror service requests to the platform, such as requesting a Red Hat Linux operating system with WebSphere, which is received via WebSEAL and sent to CFEP (Common front-end Portal) processing, the request calls the corresponding REST API, the request information package into the information flow (message) through the bus to the background cloud server (Service Provider), with the corresponding The Red Hat Linux operating system mirroring service for WebSphere is generated, and the relevant registration information is stored in the corresponding database server. So the IBM cloud computing environment includes three groups of services: A. IHS Portal Services B. CFEP Web and database Services C. Service Provider Web and database services, each with a corresponding hardware environment to support, let's look at the physical architecture of the IBM cloud computing environment.

Figure 2. IBM Cloud computing Environment Physical architecture diagram

The yellow area in this image is the front-end service configuration for the IBM cloud computing environment, and the blue area is the cloud service backend configuration. The background service configuration is the core of the whole cloud environment, is a single point fixed service unit, and front-end services can be extended across regions according to business requirements, that is, multiple service units distributed in different physical locations can be provided. For ease of parsing, the front-end services of this cloud environment are defined as two sets of hardware environments (Front end area, FE), distributed in two virtual LAN vlan1,vlan2, Background services (service Provider area referred to SP) configuration and Vlan3, Collectively known as the BSS platform.

The daily update of the deployment program in the IBM cloud computing environment is that the BSS service is distributed in the Singapore Cloud Environment Data Center, and the deployment release package file is in the United States Austin Publisher (Austin build Machine), The deployment execution environment is the automated Script Runtime Environment (single point Execution Machine) in Shanghai, China. As shown in the following illustration:

Figure 3. IBM Cloud computing Environment Automation Deployment architecture Map

The steps for deploying programs in this schema are:

from the US Austin Publisher, automatically detects and transports newly released deployment package files to the Automated scripting Environment (single point Execution Machine) at single point Execution Machine Run an Automation script to perform a deployment deployment update Front end and service Provider database on Front end and service Provider WEB Server Update deployment Dynamic application deploy static program component on IHS Server re-register Program Component Update Portal WEB file Update Info Center information repository Automation script execution completed publish deployment execution log