Publishing - Oracle SOA Suite 11gR1 Developer's Guide, Packt Publishing. Performance tuning, and capacity planning of Oracle Fusion Middleware products.

• Master Production Schedule
• Production Capacity Planning

This document provides an overview of the Oracle Fusion Applications provisioning process on the Cisco Unified Computing System ™ (Cisco UCS ™) platform. Provisioning is the act of installing, configuring, deploying, and removing Oracle Fusion Application components.

Given the number of Oracle Fusion Middleware components, as well as the number of dependencies among Oracle Fusion Application products, the goal of the provisioning framework is to limit the need to install and configure these items manually. The provisioning process mandates definition of the application topology, which includes selecting the Oracle Fusion Application you want to provision, the number of servers, the IP address and host name for each component, etc. At the end of the provisioning process, the system is operational, and an Oracle Fusion Application administrator will be able to log in to the application and begin the process of application functional configuration through the Oracle Fusion Functional Setup Manager application (beyond the scope of this document). Cisco Unified Computing System is the first truly unified data center platform that combines industry-standard, x86-architecture blade and rack servers with networking and storage access in a single system. The system uses a wire-once architecture with a self-aware, self-integrating, intelligent infrastructure that eliminates the time-consuming, manual, error-prone assembly of components into systems. Cisco UCS is built using the components illustrated in Figure 1. Cisco Unified Computing System Architecture.

Cisco UCS 6120XP 20-Port and 6140XP 40-Port Fabric Interconnects offer 20 or 40 fixed ports and one expansion module slot. These low-latency, line-rate interconnects provide all of the necessary downstream bandwidth to the solution's blade server chassis. They can connect upstream to standard Ethernet switches and storage systems. The fabric interconnects' cut-through architecture and jumbo frame capacity reduce latency and increase bandwidth to 10 Gigabit Ethernet-equipped storage systems such as EMC VNX storage systems. A fabric interconnect's capacity can be extended by populating its expansion module slot with one of four modules that increases 10 Gigabit Ethernet connectivity, connects to native Fibre Channel networks at up to 8 Gbps, or both.

Cisco UCS 6200 Series Fabric Interconnects support the system's 10-Gbps unified fabric with low-latency, lossless, cut-through switching that supports IP, storage, and management traffic using a single set of cables. The fabric interconnects offer virtual interfaces that terminate both physical and virtual connections equivalently, establishing a virtualization-aware environment in which blades, rack servers, and virtual machines are interconnected using the same mechanisms. The Cisco UCS 6248UP 48-Port Fabric Interconnect is a on rack-unit (1RU) fabric interconnect that offers up to 48 universal ports that can support 10 Gigabit Ethernet, Fibre Channel over Ethernet (FCoE), or native Fibre Channel connectivity. The Cisco UCS 6296UP packs 96 universal ports into only 2RUs. Cisco Fabric Extenders. Cisco Nexus ® 2232PP 10GE Fabric Extenders extend the unified fabric and management planes to racks for connecting rack servers to the system.

Master Production Schedule

Cisco Nexus fabric extenders provide a simple, low-cost alternative to ToR switching, supporting rack-at-a-time server deployment without adding a single management point to the system. Today, separate data and management cables connect servers to the fabric extenders within each rack, with both data and management planes carried over a single set of cables to the fabric interconnects.

Cisco server innovations prepare Cisco UCS C-Series Rack-Mount Servers for future one-wire management in which both data and management planes are carried over a single cable with racks. Cisco UCS Blade Chassis and Racks.

The Cisco UCS 5100 Series Blade Server Chassis features flexible bay configurations that can support up to eight half-width blades or up to four full-width blades in a compact 6RU form factor. The blade chassis is a highly simplified device that contrasts with traditional blade chassis, which host multiple switches and management modules. The chassis adds no points of management to the system, because it is logically part of the fabric interconnects. The Cisco UCS 5100 Series chassis hosts up to two fabric extenders, low-power-consumption devices that provide the chassis with the power budget and airflow needed to support multiple future generations blade servers and network connectivity options. The blade chassis can be configured with up to four N+1 and N+N 2500-watt (W) grid-redundant power supplies that are designed for 92 percent efficiency and to deliver high efficiency at low power draws.

Delivering performance, versatility, and density in servers designed without compromise, Cisco UCS servers power every workload, including network infrastructure applications, virtualization, cloud computing, high-performance computing (HPC), high-frequency trading, big data, and mission-critical enterprise applications and database management systems. Powered by the latest Intel Xeon processors, Cisco UCS servers form the core of a flexible and efficient data center that meets diverse business needs with a balanced combination of performance, built-in capabilities, and cost effectiveness. Cisco UCS is platform neutral, with the capability to incorporate both rack and blade servers into a single unified system.

Cisco UCS B-Series Blade Servers provide a comprehensive line of two- and four-socket servers to deliver world-record-setting performance to a wide range of workloads. Cisco offers the first server anywhere with built-in programmable I/O capacity in which the number and type of I/O devices can be configured on demand to satisfy the needs of almost any operating system or hypervisor. Cisco innovations embodied in Cisco UCS blade servers include patented Cisco Extended Memory Technology, one of the reasons that Cisco has a history of leading the market with some of the greatest memory capacity among blade servers. Among the many vendors offering servers based on Intel Xeon processors, only Cisco unleashes their full power with a balance of memory and I/O capacity that does not hold the system back.

Cisco UCS Virtual Interface Cards. Cisco UCS virtual interface cards (VICs) provide up to 256 I/O interfaces that are configured on demand to meet operating system, hypervisor, and application needs. Available in PCIe, mezzanine, and modular LAN on motherboard (mLOM) form factors, Cisco VICs can be configured with any combination of Ethernet network interface cards (NICs) and Fibre Channel host bus adapters (HBAs), carrying I/O traffic securely and with individual quality-of-service (QoS) controls directly to the system's fabric interconnects. Cisco VICs support up to 80 Gbps of network connectivity, with hardware PortChannels maintaining traffic flow over a set of active-active links. Acting as an adapter fabric extender, the cards connect fabric interconnect ports directly to operating systems and hypervisors.

With Cisco Data Center Virtual Machine Fabric Extender (VM-FEX) technology, the cards connect fabric interconnect ports directly to individual virtual machines, offering exceptional visibility and control over virtual machine networking while eliminating up to two layers of switching that other solutions require. Integrated and Embedded Management with Cisco Service Profiles.

Traditional solution architectures involve the manual assembly of components, with management as an add-on afterthought. Management servers and networks must be assembled, configured, and maintained. Blade chassis contain multiple management systems that must be wired and configured and firmware that must be kept up-to-date. Individual element managers must be used to configure system components, including RAID controller settings and firmware, BIOS firmware and settings, firmware revisions, and settings for network components including NICs, HBAs, and blade-server-resident switches.

Although some vendors claim that this scenario constitutes integrated management, if it requires a management server and manual configuration, the approach is not integrated. Cisco UCS uses true integrated, centralized, embedded management. Cisco UCS Manager resides on the Cisco UCS 6100 Series Fabric Interconnects and operates with high availability when two fabric interconnects are used, as they are in Oracle Fusion Middleware 11g.

Production Capacity Planning

Logically, every system component is aggregated into the system and managed as if it were part of the fabric interconnects. Physically, components are distributed among data center racks and blade server chassis. Cisco UCS Manager acts as an element manager that allows every component to be configured from a single pane. For organizations that want to integrate Cisco UCS with higher-level management tools, Cisco UCS Manager can export configuration information for use by ITIL processes (including configuration-management databases).

It also offers a robust XML API to facilitate deep integration with broader systems management tools. The system and its resources are stateless, and Cisco service profiles contain all the information needed to fully define and provision a server (Figure 2), including RAID levels, BIOS settings, firmware revisions and settings, adapter identities and settings, VLAN and VSAN network settings, network QoS, and data center connectivity.

Oracle Fusion Middleware 11g uses predefined, downloadable service profiles that provision each of the solution's server components, allowing the solution to be deployed rapidly and accurately in minutes, rather than the days and weeks required when all the data center subject-matter experts must use individual element managers to configure each component in their domains. Cisco Service Profiles Provide Automatic, End-to-End Configuration of the Entire Hardware Stack. Cisco service profiles provide immense benefit to the solution. They reduce the time-consuming, error-prone, manual assembly of components to a policy-based click-of-the-mouse operation. They accelerate infrastructure scaling by allowing organizations to add incremental resources and put them to work more quickly and accurately. Servers can be pooled, and service profiles can be assigned to specific slots, so that any server inserted into the slot is provisioned and put to work automatically. Service profiles enable true workload portability so that if a server upgrade is required, an existing service profile can be applied to the upgraded server and put into service within minutes.

Service profiles also reduce the number of spares that an organization needs to maintain, because a single spare can be put to any use rapidly simply by invoking the applicable service profile. Virtualization-Optimized Computing Resources. Achieving the best price-to-performance ratio in a virtual desktop environment requires an efficient, scalable, virtualization-optimized infrastructure that is optimized for the unique workload requirements of virtualized environments. Oracle Fusion Middleware 11g is implemented using Cisco B-Series Blade Servers housed in Cisco UCS 5108 Blade Server Chassis. Some of the blade servers used in the solution use Cisco Extended Memory Technology, which provides up to 384 GB of memory in a two-socket server.

The servers use Cisco VICs, which accelerate I/O in virtualized environments while increasing visibility and control over virtual machine network traffic. The blade chassis are 6RU devices that have removable partitions to allow them to contain up to eight half-width blade servers or up to four full-width blade servers. Every component except the midplane is customer replaceable. Efficient power supplies are configurable as nonredundant, N+1 redundant, or grid redundant; the chassis' eight fans are hot swappable, all cables enter from the rear of the chassis, and all blade servers are accessible from the front panel. Cisco UCS B250 M2 Extended Memory Blade Server. The Cisco UCS B250 M2 Extended Memory Blade Server increases performance and capacity for demanding virtualization and large-data-set workloads.

The server is a full-width, two-socket blade server with substantial throughput and more than twice the memory capacity of other Intel Xeon 5600 series-based two-socket servers. It takes advantage of the most advanced Samsung 40 nm DDR3 ultra low power memory technology. A Cisco UCS 5108 Blade Server Chassis can house up to four Cisco UCS B250 M2 Extended Memory Blade Servers, with a maximum of 160 per Cisco Unified Computing System. Each Cisco UCS B250 M2 uses network adapters for consolidated access to the unified fabric. This design reduces the number of adapters, cables, and access-layer switches needed for LAN and SAN connectivity. This Cisco innovation can significantly reduce capital and operating expenses, including administrative overhead, power, and cooling costs.

Network adapter choices include adapters optimized for virtualization, compatibility, and efficient, high-performance Ethernet. Cisco UCS B200 M2 Blade Server. Each Cisco UCS B-Series Blade Server uses Cisco network adapters for consolidated access to the unified fabric. The Cisco UCS B230 M1 can accommodate one mezzanine-format adapter. Oracle Fusion Middleware 11g uses the network adapters most appropriate for the customer's choice of hypervisor. Cisco UCS M81KR VICs are optimized for virtualization and provide up to 128 virtual devices, including any combination of NICs and HBAs. Note: Cisco VICs were used for desktop virtualization with VMware vSphere and EMC storage.

Virtualized I/O and Cisco Unified Fabric. Virtualized environments are I/O intensive, and best practices for software such as the VMware vSphere hypervisor require the use of separate, redundant NICs and HBAs for functions including VMware vmconsole, vmkernel, and vMotion; virtual machine traffic; and storage access.

In traditional server environments, this practice requires equipping each server with the appropriate number of interfaces and cabling each one to upstream Ethernet and Fibre Channel switches. At this point, the server is for all practical purposes dedicated to a single function for its entire lifecycle because of the time and complexity required to equip it to run different applications. Cisco UCS is stateless down to the type and number of I/O devices installed in a server, allowing just-in-time provisioning of such fundamental characteristics as I/O interfaces. This capability accelerates virtual infrastructure deployment, supports true workload portability, and extends the lifecycle of servers because they can be repurposed simply by applying a different Cisco service profile. This virtualized I/O capability is enabled by the Cisco UCS M81KR VIC, which can be configured to provide up to 128 virtual devices per server. These are standard PCI devices that appear on the server's PCI bus. The actual number of available interfaces is determined by upstream switch resources.

In the context of this solution, the card can support up to 58 virtual devices. A single Cisco VIC can thus support all of a hypervisor's I/O requirements with a single interface. It also can provide superior network throughput because it supports a separate I/O queue for each device. Oracle Fusion Middleware 11g uses Cisco VN-Link technology to provide exceptional visibility and control over network links connected to hypervisors and virtual machines. Cisco VN-Link technology makes virtual links just as visible and manageable as physical links.

Now network security, QoS, and network settings can be applied on a per-virtual-machine basis, and they remain constant regardless of virtual machine location. This capability overcomes the least-common-denominator security that is often applied in virtualized environments to facilitate virtual machine movement. Oracle Fusion Applications is standards based, making it highly adaptable. This standards-based technology enables organizations to respond effectively to change with flexible, modular, user-focused business software that is powered by best-in-class business capabilities built on open standards. Its technology framework includes the following products:.

Oracle WebCenter provides design and runtime tools for building enterprise portals, transactional websites, and social networking sites. Oracle Business Intelligence provides a full range of business intelligence capabilities that enable customers to analyze, present, report, and implement contextual actions for organization data. Oracle Universal Content Management enables customers to use document management, web content management, digital asset management; and records retention functions to build and complement their business applications. Oracle Service Oriented Architecture (SOA) Suite provides a complete set of service infrastructure components for designing, deploying, and managing SOA composite applications.

Oracle SOA Suite enables services to be created, managed, and orchestrated into SOA composite applications. Oracle WebLogic Server is a scalable, enterprise-ready application server based on Java Enterprise Edition (Java EE). Oracle JDeveloper is an integrated development environment with end-to-end support for the modeling, development, debugging, optimization, and deployment of Java applications and web services. Oracle Enterprise Manager offers business-based applications management, integrated application-to-disk management, integrated systems management, and support experience. Oracle Identity Management enables organizations to manage the end-to-end lifecycle of user identities and to secure access to enterprise resources and assets.

The Oracle Fusion Applications user interface facilitates the customer-based, intuitive design of the applications, which results in large productivity gains. The user interface design of Oracle Fusion Applications is:. Role based, enabling pervasive delivery in multiple modes, on multiple devices, and through multiple channels. Configurable and extensible, through Oracle JDeveloper during design or the composer during runtime, enhancing productivity for individual users and groups of users. Composite and contextual, providing integrated information in the context of processes. Social and collaborative, offering built-in user communities and workspace, Web 2.0 information distribution, and embedded social computing to improve collaborative work.

Oracle Fusion Applications is offered with the following deployment options:. On premises, hosted by the enterprise. Oracle cloud (available to the general public), hosted over the Internet by Oracle, software as a service (SaaS), or Oracle business partners offering business process outsourcing (BPO) solutions. On Premise cloud (available internally behind a firewall), hosted as a SaaS or BPO offering. Hybrid, as a cohesive integrated implementation both on premises and in the cloud Deployment Topology. In the topology in Figure 4, the primary node (also known as a host and shown as CRMHOST1 in the figure) is actively running the Oracle Fusion Applications instance. The secondary node (CRMHOST2) is the redundant (high-availability) node for the Oracle Fusion Applications instance.

The primary node consists of an administration server and applications that have been deployed to managed servers. Managed servers can be grouped into clusters to provide scalability and high availability for applications.

Together, the primary and secondary nodes form a domain. The solution presented here demonstrates the steps required for provisioning the Oracle Fusion HCM application module on to the Cisco UCS platform in a lab environment to illustrate the readiness of the Cisco UCS platform for provisioning Oracle Fusion Applications (all Oracle Fusion Applications products are based on a common standards-based architecture and are deployed in the same way). Please see the disclaimer about the scope and use of this document and refer to the comprehensive Oracle Fusion Applications product documentation. Following are the main provisioning goals:.

Install the applications, database, and middleware components required to run Oracle Fusion Applications. Use a guided process to gather details about the new environment using an online interview flow. The details collected form the basis for a provisioning plan that you can apply when you are ready to perform the physical installation, configuration, and deployment of the application product offerings that will run in the environment. Configure the applications, database, and middleware components and deploy them in the environment.

Cisco Unified Computing System Setup and Configuration. For a small-scale, single-node database, use the Oracle Fusion Applications Provisioning Wizard and select the Install an Applications Transaction Database option to install a single-node instance of Oracle Database Enterprise Edition. Alternatively, you can install the database manually (interactively) if you are creating a production-scale, multiple-node database. Oracle Fusion Applications also supports Oracle Real Application Clusters (RAC). The Oracle Fusion Applications software provides a repository of installers, each called silently when needed to perform application-specific tasks during the provisioning of a new environment.

The Oracle Fusion Applications Repository Creation Utility (RCU) is a GUI- and CLI-based tool used to create and manage database schemas, seed data, and PL/SQL packages required by Oracle Fusion Applications. You need to run RCU after the relational database system (RDBMS) is installed and configured, and an empty database has been created. The main features of the RCU are:. Custom schemas and tablespaces: The RCU provides the flexibility to create custom schemas and table spaces. Users can change the tablespace allocation so that components can share a single or multiple table spaces. Global and component-level prerequisites: At runtime, the RCU performs checks against both global and component-level prerequisites. If a prerequisite is not met, the RCU may post a warning and allow the procedure to continue (soft stop), or it will notify the user that a prerequisite must be met before the operation can continue (hard stop).

After the provisioning plan is generated, the next step is to provision the Applications environment on each host specified in the provisioning plan. Two tools are available to provision the Applications environment:. You can use the provisioning CLI, which provides granular control over where and what is implemented on each particular host. You can use the same wizard that generated the provisioning plan to also provision the Applications environment. This option in the wizard asks additional questions (application root, and installation library location) and allows you to make last minute updates to the offering configuration prior to implementing the plan. Designed from the start using the latest technology advances and incorporating best practices gathered from thousands of customers, the Oracle Fusion Applications suite offers completely open, service-enabled enterprise applications. Built on a foundation of 100 percent standards-based middleware and using a SOA approach and a common data model, Oracle Fusion Applications products set a new standard for the way customers innovate, work, and adopt technology.

Delivered as a complete suite of modular applications, Oracle Fusion Applications products work together with your existing application investments, helping you improve performance, lower IT costs, and get better results.