Introduction
This chapter gives an overview of the cloud computing concepts, explains the differences between IBM PureSystems offerings, and describes the fundamentals of IBM PureApplication System.
The following topics are covered in this chapter:
1.1 The cloud environment
As technology evolves, new concepts and paradigms emerge from the industry and achieve such an important position and public acceptance that they quickly become a de facto standard. With cloud computing, this fact is no different. Cloud computing is revolutionizing the way organizations provide, manage, and use IT services, and in a manner that is affecting all the connections in the information chain, from the infrastructure providers and independent software vendors (ISVs) to the users.
The term cloud computing has the following definitions, which can vary depending on the perspective where it is described:
•From the infrastructure provider perspective, cloud computing is a flexible way to deliver high-end computing resources that can be shared by multiple software vendors and scaled on demand as the workload grows.
•From the independent software vendor view, the cloud is a strategic platform that allows cost-effective business operation and rapid deployment of services, which are achieved through the optimized usage of IT resources and the immediate availability of the infrastructure.
•From the users position, the cloud is a ubiquitous environment that provides pay-per-use applications over the network that can be seamlessly integrated with multiple devices, regardless of their location.
In the next sections, we explore the characteristics, benefits, and challenges of this new cloud computing environment.
1.1.1 Characteristics
A cloud computing environment generally exhibits the following characteristics:
•Elasticity
•Resiliency
•Multitenancy
•Virtualization
•Workload migration
However, the following characteristics are essential for a solution to be referred to as a cloud:
•Resource provisioning
A real cloud environment must have an abstraction layer between the physical hardware (for example, processing, memory, storage, and network) and the computing resources that are provisioned. The virtualization infrastructure must dynamically manage and allocate these resources and support changes in the hardware capacity without compromising other resources.
•Scalable environment
Cloud environments are supposed to be elastic; that is, increase or decrease the computing capacity rapidly in response to the workload. The quality of service (QoS) must remain the same under any conditions. Computing capacity changes should be performed in a manual or automated fashion, without any disruption in the service.
•On-demand management
In a cloud environment, the customer manages the computing resources and the software capabilities on-demand, without any human intervention from the service provider. It is expected that the customers can have a single, fully integrated interface where they can manage most of the aspects of the service, but with a higher level of abstraction (for example, response time, bandwidth, and CPU usage).
•Service measurement
It is necessary for the cloud environment to provide different types of metering so that the resources consumption can be monitored, controlled, and charged back to the customer based on the service usage. This paradigm is important when there is a service level agreement (SLA) with which the customer can evaluate if the QoS is being achieved as per the contract.
•Pervasive access
Applications and services that are made available in the cloud must be pervasive; users should be able to access them from any place over the network and through standard protocols that are supported by different devices (for example, desktops, tablets, and smartphones).
1.1.2 Cloud deployment models
The cloud computing environments usually are classified according to the levels of access, control over the infrastructure, and sharing of the data center by the customers of the service (businesses or individuals). Also known as deployment models, this classification is divided into the following types:
•Public cloud
In a public cloud, the infrastructure is made available for the general public and is offered by organizations in a free or pay-per-use model. The computing resources are hosted by the service provider and shared between different customers (multi-tenancy), which have access to the cloud environment through an ordinary Internet connection.
•Private cloud
In a private cloud, the infrastructure is provisioned for an exclusive organization or department, which introduces higher levels of security and privacy into the cloud environment. The computing resources can be hosted internally or externally and managed by the organization or a third party. Usually, the access to a private cloud is restricted, with servers that are standing behind multiple firewalls and reachable only through a virtual private network (VPN).
•Community cloud
In a community cloud, the infrastructure is shared by a specific group of organizations with a common concern (for example, policy, security requirements, and compliance regulations). The computing resources can be owned, managed, and operated by one or more organizations or by a third party and the costs are spread among the members of the community.
•Hybrid cloud
A hybrid cloud is a combination of two or more cloud infrastructures (public, private or community), which offers the benefits of multiple deployment models. The computing resources are bound together by proprietary or standardized technologies, which improves the cloud fault-tolerance capabilities and enables a more flexible environment by supporting different levels of requirements.
1.1.3 Cloud service layers
Cloud computing providers can be classified according to the service layers or delivery models that the resources are offered, from the basic infrastructure services up through the platform and application services. Although there are variations on the definitions and ranges of those layers, the IT community agreed on three service models: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS).
Infrastructure as a Service
The IaaS model takes place when a cloud service provider delivers infrastructure assets, such as servers, processing, storage, network, and other fundamental physical resources to the customer. Usually, those resources are exposed through a virtualization layer, which consists of the hypervisor and its associated virtual machines (VMs), as shown in Figure 1-1.
Figure 1-1 The IaaS model and its components
The customer is responsible for installing and managing the software, which might include operating systems, middlewares, and applications. Although this approach provides the organizations a higher degree of freedom and control over the environment, the service provider is not required to offer any cloud-specific feature, such as scaling capabilities and on-demand management.
Platform as a Service
In the Platform as a Service (PaaS) model the provider delivers a complete development platform, including operating systems, runtimes, and middlewares. The objective is to supply the necessary components for rapidly building and deploying applications in the cloud, as shown in Figure 1-2.
Figure 1-2 The PaaS model and its components
The PaaS model relieves the organization of the complexities of managing individual hardware and software elements, which can be time-consuming and require highly skilled professionals to support its operation. Some providers also require that applications be written by using specific programming languages, frameworks, and application programming interfaces (APIs) that are supported by their environment. In the PaaS model, the customer does not manage the underlying infrastructure directly, but has some control over it through the operating system or some administration console that is provided with the platform.
Software as a Service
In the Software as a Service (SaaS) model, the provider delivers a fully functional application, service, or business process that runs on a cloud environment. The customer can adopt the software with little or no effort, depending on the degree of integration and adaptation that is necessary. Also, customers can collaborate with each other and share and create custom applications by composing multiple services. The basic structure of a SaaS model is shown in Figure 1-3 on page 6.
Figure 1-3 The SaaS model and its components
In a SaaS model, the software is managed in a centralized manner by the provider, which is responsible for fixing, updating, and monitoring the components of the system. The clients can configure some basic capabilities of the service, but have no control of the installed software or the underlying cloud infrastructure. The applications are accessible through a network connection from different clients, which might vary from a user’s web browser that accesses an email application to an Enterprise Service Bus (ESB) that calls a web service that is provided by a business partner.
1.1.4 Benefits of the cloud
The benefits of cloud computing are spread over all service layers. Some of these benefits are described in the next sections.
Efficient usage of the infrastructure
The cloud computing model promotes efficient usage of the infrastructure by using virtualization capabilities and multi-tenancy features. It allows multiple operating systems, middlewares, and applications to run in a single physical server. This configuration assures that the computing resources are used in an optimized and effective manner. This model also permits each tenant to run in an isolated environment, with no effect on the resources that are allocated to other tenants, and by supporting a reasonable level of security.
Dynamic and scheduled provisioning of resources
The abstraction layer that is created by virtualization allows computing resources to be dynamically allocated to the tenants. Companies can request resources manually or by establishing rules that are based on the workload so that the resources are available where and when necessary. Typically, the companies develop a workload management process that keeps priorities in check. They can also create scheduling policies; for example, to accommodate peak loads that are expected for a determined period. This ability ensures that the pool of resources is managed in the most effective manner, without waste or lack of resources.
Resiliency and workload migration
The ability to completely isolate the failure of a computing resource without affecting users is mostly related to the capability of the cloud to transparently migrate its workloads to another physical resource and continue their processing. If a compute node or storage resource fails, the cloud resiliency feature enables it to move the workload without human intervention or disruption in the service. Migration also can be started based on certain policies, such as compliance regulations (sensitive data that must be processed in a specific geographic region or data center) and performance considerations, such as user data must be processed near the user location for faster access.
Optimization of operational and capital expenses
There are several financial benefits concerning the adoption of a cloud computing solution. These benefits generally can be placed into the following categories:
•Optimization of operational expenses (OpEx) often is associated with the maintenance and operation of the IT infrastructure, including physical assets and human resources. Cloud environments enable businesses to use these assets more efficiently and at a rate that is proportional to the value that is generated from the supported application workloads. Also, the automated and dynamic nature of the cloud allows the allocation of human resources only when necessary, which leaves the employees to focus on more productive activities.
•Optimization of capital expenses (CapEx) is related to the investments and costs that are incurred for buying the IT infrastructure, which includes hardware and software assets. As the infrastructure resources can be obtained according to their needs, organizations can better manage their budget by diluting the costs over time instead of spending the money all at once. More cost savings can be derived from a higher usage of the existing hardware assets without the need for complex forecasting models, which can be inaccurate. Software licensing also can be optimized because the number of required licenses per-installation can be reduced by vertically scaling the infrastructure.
1.1.5 Challenges and requirements
While the cloud environment provides benefits, some challenges can still exist that require solutions from the cloud service providers, as described in the next sections.
Automation
In a PaaS cloud environment, it is important not only to improve the process of developing and delivering an application, but the lifecycle of that application. Some common tasks, such as resource provisioning, middleware configuration, and application deployment must be fully automated to gain efficiency, reduce the number of errors, and ensure consistency in the process. This automation makes the dynamic characteristics of a cloud possible and simplifies the application lifecycle management.
Standardization
A common concern that is shared by any new client that is adopting a cloud computing model is the lack of standards. Because cloud computing is in its infancy and most of the standardization attempts are a work in progress, it is difficult for the clients to define what solution best fits their mid to long-term needs. Data conversion and portability, application interoperability, and integration are important aspects of the technology that allow clients to deal with the heterogeneity of the cloud environments. In addition, standardization enables flexibility and vendor-neutral applications, which gives the clients freedom of choice to move from one service provider to another. Service providers must work together and collaborate with standards organizations to make sure that the cloud technology is based on open standards and delivers the necessary agility that the IT industry demands.
Security
Besides the usual challenges of developing secure IT systems, cloud computing presents another level of risk because essential services usually are outsourced to a third party and shared by multiple tenants. The external aspect of outsourcing makes it harder to maintain data integrity and privacy, support IT governance, and ensure compliance. Even basic tasks, such as applying patches and configuring firewalls, can become the responsibility of the cloud service provider. As a result, clients must establish trust relationships with their providers and define strict SLAs in terms of how these providers implement, deploy, and manage security on their behalf and what compliance levels are required.
Transaction processing
Transaction processing traditionally is one of the areas with greater demand for high-performance systems. Normally restricted to financial institutions that run on dedicated mainframe platforms, these applications are CPU and storage intensive and they are required to process thousands of transactions per second. With the increase of B2B and B2C systems, those workloads must be managed across multiple computing environments and business processes, but with the same required response-times.
Big data analytics
Managing big data sets and making them available for deep analysis are complex computing and data management challenges. Most of the existing data warehouse systems are built on old database technologies, which were not designed to scale and perform consistently for huge data volumes. To run analytic processing at top speed, general-purpose database management systems must be tuned and configured, and the queries that are manually optimized by database administrators. It is necessary that new analytic infrastructures provide an abstraction layer that hides the increasing complexity of data processing from users, which enables greater efficiency while simplicity is maintained.
1.2 IBM PureSystems solutions
To meet current IT and business demands, IBM is introducing a new category of systems. These systems combine the flexibility of general purpose systems, the elasticity of a cloud computing environment, and the simplicity of an appliance that is tuned to the workload. IBM PureSystems solutions are the building blocks of those capabilities. This new category of systems represents the collective knowledge of thousands of deployments, established guidelines, best practices, innovative thinking, and distilled expertise.
IBM PureSystems includes integration across compute, storage, networking, and platform middleware for physical and virtual resources. It also includes built-in expertise from IBM factory pre-configuration, integration, and testing. This simplified experience enables clients to deliver new cloud services and applications faster and with improved economics over traditional IT.
IBM PureSystems offerings are designed to deliver value in the following ways:
•Built-in expertise: Systems must capture and automate best practices and expertise, which reduces manual steps that affect a project’s time to value with an open architecture and allows participating solution providers to optimize their applications workloads.
•Integration by design: All hardware and software components must be integrated by design, tuned in the lab, and pre-packaged in the factory into a single, ready-to-go system that is optimized for the business task.
•Simplified experience: IT staff and the lines of business that use IT realize a simplified systems lifecycle. Collections of hardware, middleware, and application components no longer need to be separately procured, configured, tuned, and managed. IBM PureSystems are ordered, unpacked, plugged in, and managed as a single system from a centralized interface.
The IBM PureSystems offerings are available in the following types:
•IBM PureFlex™ System combines servers, storage, networking, and virtualization capabilities into a single, unified, and upgradeable infrastructure system.
•IBM PureApplication System provides a flexible platform that is designed to ease the process of deploying, customizing, and managing middlewares and applications.
•IBM PureData™ System integrates specialized hardware and software components for transactional, operational, and deep analytical workloads.
1.2.1 IBM PureFlex System
The IBM PureFlex System is an infrastructure system that provides a fully integrated computing platform. It combines servers, storage, networking, virtualization, and management into a single structure. Its built-in expertise enables organizations to manage and deploy integrated patterns of virtual and hardware resources through a unified management console. PureFlex System is ideally suited for customers who want a system that delivers the simplicity of an integrated solution while still able to tune middleware and the runtime environment.
IBM PureFlex System implements the IaaS layer and can provide higher levels of control, efficiency, and operating agility. By using built-in virtualization and workload placement that is based on resource availability, the infrastructure system enables automated scaling of resources and true workload mobility. This combination enables businesses to rapidly deploy IT services at a reduced cost. Moreover, it enables deep integration and central management through a comprehensive, open-choice infrastructure system. It dramatically reduces the skills and training that are required for managing and deploying the system.
The IBM PureFlex System is available in the following configurations:
•IBM PureFlex System Express is designed for small and medium-sized businesses and is the most affordable entry point for PureFlex System.
•IBM PureFlex System Standard is built for application servers with supporting storage and networking, and is designed to support key ISV solutions.
•IBM PureFlex System Enterprise is optimized for transactional and database systems. It has built-in redundancy for highly reliable and resilient operation to support the most critical workloads.
The PureFlex System is built from no-compromise building blocks that are based on IBM technologies that support open standards and offer roadmaps for integration with multiple systems. It is designed to give control and choice without increasing complexity. The customers are not limited to a single architecture or vendor’s middleware; they can choose from a broad range of servers, operating systems and hypervisors, and design systems that are truly optimized for their needs. Customers also can consolidate different workloads from many resource-consuming servers onto a single PureFlex System and control it all from a unified interface.
1.2.2 IBM PureApplication System
The IBM PureApplication System is a platform that is designed and tuned specifically for transactional web and enterprise applications. It provides a full stack of software, including operating systems, middlewares, databases, and patterns of expertise that enable customers to achieve greater agility and to deliver new business capabilities. IBM PureApplication System uses IBM PureFlex System infrastructure as a foundation to support high workloads, simplify infrastructure management, and improve the application’s lifecycle. This workload-aware, flexible platform is easy to deploy, customize, manage, and operate in a traditional or private cloud environment.
IBM PureApplication System implements the PaaS layer to provide superior IT economics. With the PureApplication System, customers can create their own patterns of software, middleware, and virtual resources. Customers can provision and share these patterns within a unique framework that is shaped by IT guidelines, best practices, and industry standards. These standards were gathered from many years of IBM experience with thousands of clients around the world and are infused throughout the system.
IBM PureApplication System provides the following advantages:
•Agility: By automating key processes such as operating system installation, middleware configuration and application deployment, PureApplication System can reduce the cost and time that is required to manage the platform.
•Efficiency: With PureApplication built-in expertise, customers can optimize critical business processes and conserve valuable resources, getting the most from the systems in terms of energy efficiency, maintenance, and fast response to problems.
•Simplicity: PureApplication built-in patterns of expertise can help customers to easily consolidate different servers, storage, and applications into an easy-to-manage, integrated system that can be controlled from a single management console.
•Scalability: By defining a high-level set of parameters and policies, administrators can take advantage of PureApplication System features to scale the application up and down automatically according to the workload.
•Reliability: Built-in application expertise reduces the number of unplanned outages through guidelines and the automation of manual processes that are identified as causes of outages. PureApplication System also can use data replication to increase high availability.
1.2.3 IBM PureData System
The IBM PureData System is a platform that is optimized exclusively for delivering data services to today’s demanding applications. It provides data management expertise for each type of workload, such as automated pattern-based deployment and management of highly reliable and scalable transactional database services. It contains hardware and software capabilities that are designed and optimized for specific high-performance data workloads, such as data filtering by using programmable hardware (FPGAs) for ultrafast execution of analytic queries. The IBM PureData System provides a fully integrated management console for the entire system, with integrated system upgrades and maintenance.
Different applications rely on different types of data processing, as shown in the following examples:
•E-commerce: Relies on scalable relational databases for transaction processing.
•Customer analysis: Requires analytics data warehouses for reporting and analytics.
•Fraud detection: Needs operational data warehouses for real-time decision making.
Traditionally, IT staff are responsible for configuring and tuning general purpose systems according to each of those needs. This general purpose approach presents the following disadvantages:
•Configurations can be complex and rely on specialized skills, which may not be available.
•Configuration and tuning is error prone because it is a manual process.
•It is a time-consuming task that can affect the application time-to-market.
In response to these disadvantages, IBM adopted a smarter approach by designing, optimizing, and tuning IBM PureData System for each type of workload to deliver specialized data services for applications with simplicity, speed, and lower cost. As such, the following advantages are realized:
•Reduced complexity: Each system is pre-tuned for the different needs of different analytic and transaction processing.
•Accelerated time-to-value: Data management expertise is built in each system and ready for immediate use.
•Improved IT economics: The entire system lifecycle is simplified, from acquisition to maintenance and upgrade.
The IBM PureData System is available in the following configurations:
•IBM PureData System for Transactions: Contains factory-integrated and optimized server, storage, network, and software resources that are selected specifically for online transactional processing (OLTP) workloads. They are designed, integrated, and tuned to support transactional processing applications that require high performance, high scalability, and high throughput with fast response time.
•IBM PureData System for Analytics: Designed specifically for running complex analytics on large data volumes. This system, which is powered by Netezza® technology, is a high-performance, hardware-accelerated, massively parallel system that enables clients to perform analytics on big data. It provides an easy-to-use data warehouse appliance that runs business intelligence (BI) algorithms extremely fast and requires minimal administration or tuning.
•IBM PureData System for Operational Analytics: Optimized for the right balance of analytics performance and operational query throughput, this PureData System delivers rapid insight into high volumes of fast-moving data. It was designed for applications that must analyze multiple business variables to perform real-time decision making. By using multidimensional cubing services over a relational data warehouse schema, it optimizes performance for online analytical processing (OLAP) queries, providing more power for users to analyze data.
1.2.4 Patterns of expertise
IBM PureSystems are built to include patterns of expertise. Patterns of expertise accelerate the time-to-value of applications, services, and business processes by encapsulating guidelines and best practices into a repeatable and deployable form. They can automatically configure, manage, and optimize the elements of a solution, from the infrastructure resources up through the middleware and applications.
Patterns of expertise represent the knowledge and experience that is gained over decades by IBM while optimizing the deployment and management of thousands of data centers, software infrastructures, and applications around the world. They are categorized in three types, one for each service layer: infrastructure patterns, platform patterns, and application patterns. Figure 1-4 shows the relationship between the patterns and IBM PureSystems products.
Figure 1-4 Patterns of expertise and IBM PureSystems products
Infrastructure patterns
Infrastructure patterns provide an integrated, automated, policy-driven infrastructure management across processing, storage, and networking resources. These patterns enable a faster hardware configuration and simplified management of low-level resources, which in turn reduces operational expenses and increases performance by using optimal settings.
Platform patterns
Platform patterns encapsulate pre-configured, policy managed platform services, such as caching elasticity, failover, load balancing, security, database, and middleware. These patterns enable faster application development via the automation of common platform management tasks, such as environment setup, middleware configuration, and application deployment.
Application patterns
Application patterns comprise predefined application architectures and required platform services that are deployed and managed by the system according to a set of policies. Application patterns encapsulate proven, well-designed solutions and best practices for recurring problems in enterprise architectures. They provide key software engineering solutions for building applications that are robust, scalable, and easy to maintain.
1.3 IBM PureApplication System overview
The IBM PureApplication System is a workload optimized and integrated hardware and software solution that is designed to simplify the development, provisioning, and management of applications in a private cloud environment. It features integrated management capabilities, which allows self-service provisioning of elastic applications, databases, and middlewares. In PureApplication System, the hardware and software are deeply integrated, which provides a high degree of automation, performance, and simplicity to the data centers.
PureApplication System is integrated by design. IBM took the collective knowledge of decades of experience in integrating and tuning servers, storage, networking, virtualization, and management and used it to design a new system. With PureApplication System, the middleware, development, and deployment expertise are integrated and optimized from the factory, as shown in Figure 1-5. PureApplication System inherits the infrastructure capabilities of PureFlex System and builds on them a complete platform, which is designed for enterprise applications that require high performance, scalability, and optimal usage of resources.
Figure 1-5 PureApplication System features are integrated by design
1.3.1 Configurations
There are four available configurations of PureApplication System, as shown in Table 1-1. All four configurations include the same storage and networking components and same middleware patterns and management capabilities. Each configuration includes 48 terabytes of hard disk storage and 6.4 terabytes of solid-state storage. The configurations vary by the number of included compute nodes. Each compute node consists of 16 physical cores and 256 gigabytes of memory. The smallest configuration includes 6 compute nodes and the largest includes 38. Upgrades to a larger system can be ordered and installed without any downtime for existing workloads.
Table 1-1 IBM PureApplication System configurations
|
|
IBM PureApplication System W1500-96
|
IBM PureApplication System W1500-192
|
IBM PureApplication System W1500-384
|
IBM PureApplication System W1500-608
|
|
Compute nodes
|
6
|
12
|
24
|
38
|
|
CPU cores
|
96
|
192
|
384
|
608
|
|
RAM memory
|
1.5 TB
|
3.1 TB
|
6.1 TB
|
9.7 TB
|
|
SSD storage
|
6.4 TB
|
|||
|
HDD storage
|
48 TB
|
|||
The IBM PureApplication System comes with a set of preinstalled software that can be used in the system, including the operating system (Red Hat Enterprise Linux), middleware (WebSphere Application Server), and database (DB2 Enterprise). Other software products also are built and optimized for use in PureApplication System but are not included with the system. These products can be purchased and loaded onto PureApplication System through the PureSystems Centre, a repository of products that are built for the PureSystems family. PureApplication System includes license management capabilities so that the customers can track license usage and availability for non-entitled software products.
The system also includes built-in expertise that is provided through virtual system patterns and virtual application patterns, which enable applications to easily integrate and optimize the use of the underlying platform. It also comes with the Advanced Middleware Configuration tool for application on-boarding and the Image Construction and Composition Tool for building custom virtual images for use with PureApplication System.
1.3.2 Virtual patterns
IBM uniquely builds expertise into PureApplication System with patterns of expertise. Patterns of expertise are proven best practices and expertise for complex tasks that are learned from decades of client and partner engagements that are captured, tested, optimized, and then built into the system. They are not just a blueprint or set of instructions; the expertise is built in and inherently usable.
Patterns deliver faster time-to-value by removing manual steps and automating delivery. They improve efficiency and simplicity by reducing costs, resources, and the amount of in-house expertise that is required for deploying solutions. Patterns also increase control by allowing repeatable, automated, and optimized deployments, which reduce the risk of human error.
PureApplication System contains three types of patterns: virtual system patterns, virtual application patterns, and virtual appliances. These patterns are described in the following sections. For more information, see Chapter 4, “PureApplication Deployment Models” on page 107.
Virtual System Patterns
Virtual system patterns provide an automated model for deploying middleware topology patterns. They allow customers to quickly deploy traditional workloads in a virtualized environment in a repeatable fashion. Products that are deployed by using the virtual system patterns are managed by using the existing management tools that are provided by those products.
Virtual system patterns contain a collection of middleware parts that can be connected to build a topology for a particular type of deployment. The middleware parts that are used to build these patterns are known as virtual images. IBM provides many virtual images that contain IBM middleware products that are designed to run in virtual machine environments. It also is possible to include custom virtual images.
Virtual Applications Patterns
Virtual application patterns provide a highly automated, policy-based deployment model in which the customer defines application components and policies that specify the needs of the application. The virtual application pattern is application-centric, whereas the virtual system pattern is middleware topology-centric. It has a highly simplified administrative model, which exposes fewer administrative functions than the virtual system pattern.
IBM PureApplication System support the most common transactional web application types and includes several preinstalled virtual application patterns. After the application requirements are defined, such as what services are used and the QoS that should be applied, PureApplication System automatically deploys and configures the appropriate middleware components to run the application with optimal performance. All virtual application patterns support integrated monitoring, scaling, and failure recovery capabilities.
Virtual Appliances
PureApplication System also supports the virtual appliance deployment model, which allows the system to run custom software images that were previously modified. However, virtual appliances do not feature the robust management and monitoring features that are available for virtual systems and virtual applications. These features are useful when the pattern requires a software product from another vendor, or corporate standards require that all deployed instances contain a particular operating system component.
The following methods are supported to create custom images:
•The extend and capture function allows customizing a single image and then saving into the catalog as a new image.
•The Image Construction and Composition Tool allows the customer to build custom images for the operating system and middlewares.
1.3.3 Management
The IBM PureApplication System provides a single unified interface for managing the entire system. The integrated console provides management and monitoring interfaces for system hardware, virtualized storage and networking, license usage monitoring, user auditing, and security configuration. All of these system management functions are provided through the same interface that is used for monitoring workloads.
Monitoring
PureApplication System includes pre-configured capabilities for monitoring all the hardware and software components that are provided with the system. Workload monitoring capabilities include hypervisors, operating systems, and entitled middleware and database products. It also supports more workload types for non-included products through expanded monitoring capabilities.
Maintenance
PureApplication System maintenance is divided into the following categories:
•System maintenance: Includes updates for components that are part of the system, such as hardware and firmware components, compute nodes, switches, storage, hypervisors, and management software.
•Workload maintenance: Include updates to everything that is contained within the virtual machines that run workloads on the system, such as operating systems, middlewares, databases, and updates to virtual system and virtual application patterns.
These maintenance types can be applied independently because they are likely to be applied by different teams and on different schedules.
Licensing
PureApplication System also includes tools for managing software license usage within the system. License usage can be monitored on a per-server basis or by using the processor value unit (PVU) model, depending on which type of software license was purchased. PureApplication System tracks the licenses that are used for that product in real time as virtual machines are created and destroyed so that individual users do not have to know how many licenses are in use by other users. In addition to monitoring and optionally enforcing license usage limits, PureApplication System supports exporting historical license usage information for use in spreadsheets and other tools for analyzing license usage over time.
1.3.4 Problem determination
IBM PureApplication System includes several capabilities to assist with troubleshooting. The system logs track actions that are performed by the components and enables the administrators access to detailed operational information that is contained within the kernel, error, and trace files. It also provides information regarding the addition, modification, and deletion of auditable objects, such as users, virtual systems, patterns, and other items.
The user also has access to all of the logs that are created by the workloads, such as WebSphere Application Server, IBM HTTP Server, and DB2 logs. In addition to the log data, the system collects and notifies the user of events that are generated by the system hardware and software. They are displayed in the event notification panel in the system console, with which you can view events by severity, type, status, and time, among other criteria.
The integrated console also provides a tool that is called infrastructure map, which provides a graphical view of the system hardware that makes it easy to identify possible issues and drill down to gather more information. It displays data for all of the hardware components in the system, including overall status, graphical representations of hardware LEDs, usage, temperature, and performance data, as shown in Figure 1-6.
Figure 1-6 The hardware infrastructure map in PureApplication System console
The entire system and basic status are displayed by default in the infrastructure map. If there are errors or warnings for any of the hardware components, they are shown directly in this view. Clicking any component in the map displays more detailed information about the expansion unit as a whole, such as errors, warnings, and usage statistics.
For more information about problem determination and troubleshooting, see Chapter 8, “Troubleshooting PureApplication System Environment” on page 329.
1.4 Scope of this book
The goal of this book is to introduce new users to the concepts of IBM PureApplication System and describe the most common problems, solutions, best practices, and advanced use cases around adopting IBM PureApplication System V1.0.
The following topics are addressed in this book:
•Chapter 1, “Introduction” on page 1 provides a brief overview of the cloud computing concepts, explains the differences between IBM PureSystems offerings, and describes the fundamentals of IBM PureApplication System.
•Chapter 2, “Integrating IBM PureApplication System into an existing data center” on page 21 describes the necessary steps to integrate an existing environment with PureApplication System. It shows the prerequisites, main interfaces, and connection options.
•Chapter 3, “Usage patterns for isolating applications” on page 61 describes how applications that are deployed into a cloud environment can run independently from each other. It also explains how isolation separates application traffic through the network and prevents cross application impact that is caused by resource contention.
•Chapter 4, “PureApplication Deployment Models” on page 107 provides a review of the deployment models that are available in PureApplication System, explains their main features, implementation strategies, topologies, and tools for each model.
•Chapter 5, “Customizing Virtual System Patterns” on page 199 describes the various resources and tools that are available for customizing virtual system patterns, managing virtual images, script packages, add-ons, and configuring advanced options in PureApplication System.
•Chapter 6, “Customizing Virtual Application Patterns” on page 267 describes the various interfaces available for customizing virtual application patterns, such as extending existing out-of-box patterns or developing plug-ins.
•Chapter 7, “Integrating PureData for Transaction” on page 309 describes the advantages, prerequisites, and sample scenarios about how to use PureData System with PureApplication System.
•Chapter 8, “Troubleshooting PureApplication System Environment” on page 329 provides helpful information about the tooling, logging, and debugging features that are available in IBM PureApplication System for problem determination and troubleshooting.
•Chapter 9, “High Availability and Disaster Recovery” on page 377 covers the main failure scenarios for the different deployment models in PureApplication System to achieve high availability and perform disaster recovery.
1.4.1 Intended audience
The target audience for this book is anyone from the IT industry, including technical consultants, business partners and independent software vendors, who are considering migrating to a cloud computing solution or want to acquire a better understanding of IBM PureApplication System.
In addition, the audience includes system administrators, middleware specialists, and software engineers who seek a more in-depth exposure to PureApplication System features and capabilities.
1.4.2 Assumptions
This book makes the following assumptions to simplify the information and make it useful to as many customers as possible:
•The user has access to a deployed environment of IBM PureApplication System v1.0.
•The user is entitled to access the system and is assigned the proper roles and privileges to complete the required tasks.
•The user is familiar with the following information:
– Hardware concepts: Computing, storage, memory, networking
– Virtualization terminology: VMware ESX, hypervisor, virtual machine, image
– Networking concepts: TCP/IP, DNS, NTP, LAN, VLAN, VPN, proxy, gateway
– Security concepts: SSL, SSH, LDAP, user, group, role
– Operating systems: UNIX, Red Hat Enterprise Linux (RHEL)
– Middlewares: WebSphere Application Server, wsadmin
– Web servers: IBM HTTP Server (IHS)
– Enterprise applications: Java Platform, Enterprise Edition; JVM; EJB; EAR; WAR
– Databases: DB2 Enterprise, SQL, DDL
– Scripting languages and notations: Shell script, Python, Jython, JSON
•The user has access to the following software and tools that are required for completing the tasks that are described in this book:
– Virtual Pattern Development Kit (VPDK)
– Plug-in Development Kit (PDK)
– Command-line interface (CLI)
– Image Construction and Composition Tool
– Advanced Middleware Configuration (AMC)
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.