数据中心现代化 CON 1725 甲骨文全球大会 2013. 中国上海 朱午光 甲骨文公司资深销售顾问

以下内容旨在概述产品的总体发展方向。该内容仅供参考，不可纳入任何合同。本演示不承诺提供任何材料、代码或功能，也不应将其作为购买决策的依据。所描述的有关 Oracle 产品的任何特性或功能的开发、发布和时间安排均由 Oracle 自行决定。 Before we get started, I need to share that this webcast is for informational purposes only and is covered by this guidance.

数据中心现代化 趋势 优势 现代化 优化 简化 加速 客户 总结

CIO 优先考虑事项 着眼现在，放眼未来 “CIO 们预见到 2013 年 IT 将发生翻天覆地的变化 — 一个基于 创新、竞争优势和客户增长的 IT。CIO 们过去就有类似的雄心壮 志，但经济、战略和技术变革目前正在为几年内实现该梦想奠定 基础。” — 领导变革时代：2010 CIO 议事日程 Gartner, Inc. 2010 年 1 月 As many of you already know, 2010 is looking more and more like an important turning point in both the Global Economy and the future of the IT Industry. As the consumers and businesses cautiously resume spending and economic growth returns, CIOs and IT managers must strike a careful balance between cutting costs, improving productivitiy, and investing for future growth. Now more than ever, CIOs are faced with complex and difficult decisions that will impact the efficiency of their IT resources and the impact they will have on the overall business. In a survey of over 1500 CIOs, Gartner found that a majority were making the transition from surviving the current downturn, to planning and investing for future growth in 2013 and beyond.

外部市场转变 与技术进步 移动计算 工作的演变 大数据 云孤岛 101010 01010010 101100111 110101010 101111110 博客 智能 量表 101 01010010 10110100111 外部市场转变 与技术进步 At the same time there are external factors putting pressure on the Datacenter Infrastructure that brings Complexity. Mobile computing brings its own set of challenges to the enterprise. Everyone is accustomed to anytime, anywhere access to applications on their personal smart phones, PDAs, and tablets—and that means employees are demanding that I.T. support those devices. And by 2014, 90 percent of organizations will support corporate applications on personal devices. * (Notes: *Source: Gartner’s Top Predictions: 2011 To compete in the new global economy, companies will need to empower people to drive innovation and change. This requires a strategy for innovation work that identifies the types of information employees need to do their jobs, as well as ways in which they engage with their peers. Research shows that innovation worker productivity can increase by 50% when companies apply technology more precisely, with solutions tailored to specific jobs or roles, such as technologies like portals and collaboration tools (Notes: “Rethinking knowledge work: A strategic approach”, McKinsey Quarterly, February 2011) The unintended consequences of recent technology advances can also exacerbate complexity. The Internet generates vast amounts of information every second, leading to Big Data which describes very large datasets that can exceed an organization’s capacity – using conventional technology – to capture, store, search, share and analyze. Finally, niche SaaS solutions are recreating application silos, but this time in the cloud. Only four percent of organizations have fully integrated their SaaS silos, and 40 percent cite technology limitations as a reason for not integrating them. (Notes: “Appirio Shatters Cloud Silos With Cloud Broker Technology,” CRN, September 15, 2010) 移动计算 工作的演变 大数据 云孤岛

通过简化 IT 提升实际价值 20% 提高运营效率 创新 维护 IDC 报告称，解决方案的复杂性在很大程度上促使 IT 通常将 80% 的 IT 资源投入在维护上，而在创新上的投入仅占 20%。 20% According to an IDC report IT focuses 80% their resources on the ongoing IT Operations and Infrastructure of the Data Center as opposed to much need innovation about how/what the services created, deployed to help the business stay ahead of the competition. Simplification of IT drives creation of business value. 维护 信息来源：http://www.slideshare.net/IntergenNZ/an-update-on-idcs-top-10-predictions-for-2008-presentation

数据中心趋势 物理 专用 异构 手动管理 IT 管理 组件组装 虚拟、云 共享 标准化 自动化管理 自助服务 预先优化的系统 As you look at the transitions that are happing in datacenters, we see opportunities to achieve better than incremental improvements in several key areas. Virtual / Cloud: move past the typical infrastructure projects to the platform and application level. Pools of application resources, not just pools of compute, storage or network. Management: manage from the application down as well as the infrastructure up in an integrated way to take the mystery out of problem resolution. 2 big trends and focus areas for Oracle: - Integrated, workload optimized systems from apps to disk - Elastic cloud services, not just at the infrastructure layer, but at the platform level

简化数据中心 更高的敏捷性、响应能力 更高的 QoS、QoE 降低成本 提高利用率 打造灵活的基础架构 动态 灵活的 云 采用 虚拟化 静态 独立式 服务器 更高的敏捷性、响应能力 更高的 QoS、QoE 降低成本 提高利用率 Just as complexity increases, solutions also emerge. We believe that the enterprises are on a JOURNEY to cloud computing. Some or most will EVOLVE their current IT infrastructure to become more “cloud-like”, although at different rates, to provide greater agility and responsiveness to business needs, higher quality of service and quality of experience in terms of latency, availability, mobility, and lower costs and greater utilization. The goal is to simplify the datacenter In many cases, the technical building blocks [for cloud computing] are available, ahead of enterprise readiness, so we think that enterprises will evolve towards the right at different rates. Virtualization, together with Consolidation, is an important step for simplification.

Gartner：合理化和虚拟化 合理化 整合负载 更换旧服务器 标准化服务器 虚拟化 虚化服务器与物理服务器的比值为 10:1 200 > 170 台服务器 整合负载 更换旧服务器 标准化服务器 虚拟化 170 > 17 台服务器 虚化服务器与物理服务器的比值为 10:1 节省 32% 节省 55% 硬件 软件 工作人员 设施/功耗 合理化 之后 之前 虚拟化

提高服务器利用率 提升能力 虚拟化环境的优势 提高灵活性 减少 占用空间 Increased server utilization Lower acquisition costs Reduced service costs Reduced power and cooling costs Improved capacity and response time Newer and faster processors, interconnects, and I/O Smaller foot print Reduced data center infrastructure costs Increased flexibility Faster time to deployment of new applications There are numerous benefits from Server consolidation. The obvious benefits are those gained by improving system utilization….reduced power, space and cost and better ROI. However, there are other benefits as well. When consolidating older applications from older servers the new platforms normally provide increased performance, better virtualization, better manageability and better RAS. An environment where servers are treated a resource pools that can be partitioned to provide resources to various applications also means faster deployment of newer applications. It is easier to reassign resources inside a server to a new application than it is to deploy a new server. Note that in many cases “consolidation” means replacing older servers with newer servers. However, new applications can be deployed in a consolidated environment from the start thereby gaining most of the benefits articulated on this slide.

这不仅仅是虚拟化…… 优化：通过软件芯片化 各层内部 集成 各层之间 集成 与其他系统 集成 Slide Transition: When we say that Oracle hardware and software is engineered to work together, this refers to our unique ability to offer customers a complete hardware and software stack, from applications through middleware and databases, and all the way down into servers and storage. Integration Inside Each Layer Within each layer of the Oracle technology stack, all components are engineered to work together—either individually or as part of a product family or suite. Oracle tests everything within the layer, including interactions between the applications, middleware, and database. For example, thousands of developers spent millions of hours developing code, stress testing, cluster testing, and regression testing all the components in Oracle Fusion Middleware 11g. This level of investment ensures that all Oracle products—in every product family and every product suite—can interoperate with each another. Integration Between Layers Oracle integrates all layers of the technology stack—from the applications down to servers and storage. Every layer is tested and certified to work with every other layer: applications are tested with middleware, middleware is tested with databases, and databases are tested with servers and storage. Oracle certifies the complete stack, so customers know which versions of software are designed to work together. For example, every major application product line that Oracle delivers—including Oracle E-Business Suite, PeopleSoft, JD Edwards EnterpriseOne, JD Edwards World, Siebel CRM, and Oracle Fusion Applications—is certified with Oracle Fusion Middleware. Customers running other Oracle Applications can also take advantage of Oracle’s common set of technologies (such as business intelligence, security and identity management, and the latest user experience tools) through Oracle Fusion Middleware. Oracle packages these technologies into standardized solutions that customers can deploy and upgrade together. Because Oracle owns the entire technology stack, and because Oracle engineering teams work together closely, Oracle can optimize the entire stack’s best-of-breed components for performance, scalability, security, business intelligence and manageability, and can share these technology advances throughout the stack. Integration With Other Systems Although you get the greatest benefit from running a technology stack that uses Oracle products for all the layers, most Oracle customers have heterogeneous environments that include legacy and third party products. No problem. Each layer in the Oracle stack—and each product within each layer—is engineered according to open industry standards, so customers can integrate Oracle technology into their existing IT landscape and still realize significant benefits from using Oracle products anywhere in their technology stack. Oracle is investing over $4 billion per year in R&D to deliver best of breed and engineered systems and to provide complete end to end performance and value. We are able to deliver greater innovation more quickly than our competitors, because we coordinate development across all of our engineering teams responsible for different layers of our stack.   各层内部 集成 各层之间 集成 与其他系统 集成

Oracle 的独特优势 集成设计、卓越性能，提供前所未有的价值 加快业务流程 | 降低运营和资本支出 虚拟化 可扩展 灵活 云就绪 零开销 自动化 安全性 系统管理 集中化 深度集成 全体系支持 单点 正常运行 更快速地解决问题 EM Ops Center Oracle Solaris Zones Oracle VM Oracle 虚拟网络服务器（虚拟机管理程序） EM Cloud Control Oracle Solaris Oracle 虚拟网络服务器 存储 EM Ops Center Oracle Solaris 服务器 (ILOM) 标准支持服务 加快业务流程 | 降低运营和资本支出

优化 I’d like to close by providing a few examples of customers we’re working with.

我们的战略旨在实现数据中心加速 最佳性能。最大价值。内置云 Oracle 集成系统 计算与存储构建块 软件

Oracle Enterprise Manager 云体系管理 业务驱动的 应用管理 应用感知的 基础架构管理 管理软件 Oracle 行业应用 ISV 应用 Oracle 横向应用 定制应用 SOA Suite 和 BPM Suite 安全性： 身份管理 Data Integrator 和 GoldenGate 用户参与： WebCenter Oracle Database WebLogic 中间件和数据库基础架构 Oracle VM for x86 Oracle Linux Oracle Solaris Oracle VM for SPARC Solaris 区域 系统基础架构 网络结构 SPARC 服务器 x86 服务器 SAN 存储 NAS 存储 磁带 集成系统 THIS IS ANIMATED BUILD SLIDE WITH TWO BUILDS. We are not going to deep dive into all the Oracle products in this presentation. When you’re ready, we can bring in the individual product expertise, but it’s important to understand a little bit more about Oracle’s portfolio behind the big “red stack” graphic. We group these into three basic categories here starting with Applications. This includes core business apps like Financials, Procurement, HR with multiple product portfolios like E-Business Suite and JD Edwards. Most of these applications have some specialized options for different industries and we also have some apps designed for very specific industry needs like retail or banking or automotive manufacturing. We also include other SW vendor applications and custom applications here because the application layer is a highly specialized place. This is where many companies actually have core intellectual property – where they’re using IT to address their market needs so we want to make sure that’s captured here. CLICK: Then we have the middleware and database. Oracle has the industry leading application server and database with WebLogic and Oracle Database but we’re talking about the full middleware suite here which also includes things like SOA for application integration and identity management. And then we have the Systems Infrastructure which also includes management capabilities across the stack. And you will notice we don’t just have one hardware stack. We have a dual OS strategy with Oracle Solaris and Oracle Linux as well as virtualization capabilities with Oracle VM and Solaris Zones. We have servers that use Oracle’s own SPARC processors as well as Intel x86 processors. If you follow Oracle news, you’ve also noticed that network fabric is an area where Oracle is investing more in lately with increased R&D as well as some strategic acquisitions like Xsigo late last year. The network is really the glue that brings this all together and so we want to engineer very tightly together with the server and storage infrastructure. On storage front, we have fiber channel SAN storage and the ZFS Storage appliance for NAS environments. And of course tape systems for long term archive. And then finally the Engineered Systems. As I mentioned, Engineered Systems are really purpose built machines that combine key hardware and software elements and allow us to do some unique things to address very specific IT needs.

Oracle SPARC 服务器系列产品 企业服务器 任务关键型计算的基础 旨在提供同类最佳的性能、 可靠性、可用性和安全性 T4-4 SPARC SuperCluster SPARC M5-32 新款 SPARC T5-8 The Data Center teams are looking for opportunities to make most out of their existing infrastructure by driving up utilization rates with virtualization. As we have outlined earlier there are 3 core focus areas that need be managed efficiently for an Optimized Data Center. Efficiency through greater utilization of resources Achieving Scale reliably and securely to respond to fluctuating demands And, automation of operations in order to simplify the routine IT tasks. So ultimately what we exactly want to achieve is a Data Center that can provide services in a complete, integrated and business-driven way with minimal over-head and is primed for cloud deployments. Oracle’s unique Data Center management products and tools can help greatly towards the goal of Optimization. 新款 SPARC T5-4 新款 SPARC T5-2 T4-4 T4-2 T4-1 新款 SPARC T5-1B T4-1B

SPARC 针对全体系的优化 关键线程 2GB 大页面 表空间加密 JVM 加密 JVM 2GB 大页面 ZFS 加密 迁移加密 SPARC is more than just a fast processor. It accelerates functionality across multiple layers of software critical to a corporations operation. CRITICAL THREADS: Critical Threads optimizes performance by assigning a process exclusive access to all of a core’s hardware resources. Proves up to 30% improvement in Oracle Database and Oracle Coherence performance and up to 2X improvement for Java. LARGE PAGES Large Pages provide a faster method for applications that actually use large amount of memory to quickly find where they’ve placed their data in memory. Performance can improve up to 30% in some cases.

Oracle Solaris 针对全体系的优化 JVM控制台与Dtrace集成 RAC 内核模式加速 快速 DB 启动/关闭 动态 SGA DB DTrace 使用 ZFS 的 PDB 克隆 JVM DTrace SPARC isn’t the only Oracle technology to help accelerate the Oracle stack. Oracle Solaris 11.1 has introduced a number of features accelerate and improve system operations. RAC: Delivers more consistent and predictable RAC performance by enabling faster locks. Up to 30% lower latency and up to 20% higher throughput.

简化 I’d like to close by providing a few examples of customers we’re working with.

21 世纪的云基础架构 适用于所有任务关键型应用 x86 SPARC Oracle VM Oracle Solaris 11 Solaris 11 Zone Solaris 10 Zone Solaris 老式 Zone Some datacenter customers have mission-critical apps that are only supported on S8 and S9. While Oracle does guarantee (via our binary compatibility legal guarantee) that S8 and S9 apps will run on S10 that does not mean that an ISV will support their app on S10. Today’s SPARC servers only run on Solaris 10. So, this problem, the lack of S10 support for some applications, might be a big problem if was not for S8 an S9 Containers. This technology creates virtual S8 and S9 OS environments that can be used to run S8 and S9 applications unchanged. This protects our customers investments in applications that run only on older versions of Solaris. This allows the customer to take advantage of upgrading to newer servers but still run their S8/S9 applications. Rapid conversion from older environments Simplifies move to newer, more efficient SPARC systems Lower power/cooling/space/support costs Physical-to-Virtual (P2V) consolidation

3 3 利用 Solaris 简化云部署 IPS 信息库 IPS 信息库 产品包 产品包 打造企业云 虚拟环境蔓延膨胀 重装效率低下 缩短维护时间 高技能补丁 有风险的更新 通过 自动安装程序 实现快速供应 AI 服务器 DHCP 服务器 网络配置 AI 地址 活动 BE 活动 BE 新 BE 旧 BE 更新的 BE 慢速、 即席供应 快速开发、测试和 部署 Zone 虚拟环境蔓延膨胀 重装效率低下 Solaris simplifies the the Data Center and Cloud management with the latest tools added to Solaris 11. Jointly, OpsCenter, Solaris and SPARC servers provide a number management facilities. Here is a highlight as some of which are shown on the slide here: Automated Installer Deploy Default Manifests within the same Ops Center deployment plan interface IPS Packaging Leverage Ops Center to control a local IPS MSR and gain insight to changes via reports and agentless IPS management and provisioning SMF Insight into service awareness over just process awareness. Ability to see the relationship and control the relationship of operating system daemons and connection/utility points. Connect services to PIDs and enter OS Analytics Active Boot Environment and ZFS Control ABE via deployment plan policies. Leverage ZFS snapshots of BE and turn them into active boot environments NIC Virtualization Manage bandwidth and priority control over NICs Dtrace | FMA Explorer process performance through OS Analytics while subscribing to FMA events. IPS 信息库 IPS 信息库 3 3 产品包 虚拟机 虚拟机 虚拟机 产品包 虚拟机 虚拟机 虚拟机

显著提高生命周期管理速度 在 6 分钟内恢复服务！ 支持云 SLA 6:00 开始更新 新 安全 补丁 维护窗口：下午 6:00 - 7:00 6:00 开始更新 6:00 - 6:02 依赖性检查、补丁/更新计划 System update times may vary depending on size of systems to boot. Solaris 11 can update a system very quickly and then Fast Reboot for SPARC and x86 gets the new environment back online quickly. If you are not happy with the updates then boot the previous ZFS boot environment – Fool-proof updates Solaris 11 dramatically improves how the patches are applied and managed while shortening the systems downtime to Mission Critical Data Center requirements. 6:04 - 6:06 重启并正常运行 6:02 - 6:04 创建新引导环境，下载更新并启用 在 6 分钟内恢复服务！

Oracle VM Server for SPARC 现在适用于所有 SPARC 服务器，包括 SPARC M5-32 和 SPARC T4/T5 隔离操作系统和应用 在每个逻辑（或虚拟）域中 Oracle Solaris 10 Oracle Solaris 10 Oracle Solaris 10 Oracle Solaris 10 针对 SPARC 和 Oracle Solaris 进行了优化 针对 SPARC 和 Oracle Solaris 进行了优 化 Oracle Solaris 11 Oracle Solaris 11 Oracle Solaris 11 Oracle Solaris 11 基于固件的虚拟机管理程序 硬件级虚拟化提供逻辑域 SPARC T 系列 SPARC 虚拟机管理程序 Oracle VM Server for SPARC, previously called Sun Logical Domains, provides highly efficient, enterprise-class virtualization capabilities for Oracle’s SPARC T-Series servers. Oracle VM Server for SPARC leverages the built-in hypervisor to subdivide system resources (CPUs, memory, network, and storage) by creating partitions called logical (or virtual) domains. Each logical domain can run an independent operating system. Oracle VM Server for SPARC provides the flexibility to deploy multiple Oracle Solaris operating systems simultaneously on a single platform. This is the virtualization solution that fully optimizes Oracle Solaris and SPARC for your enterprise server workloads. 逻辑域 在专用 CPU 线程中运行 24

轻松迁移，提高可用性 通过 OVM SPARC 和 OVM x86 进行在线迁移 从之前的版本进行无缝升级 Oracle Solaris 11 Zone、Oracle VM for SPARC v2v Solaris 11 S10 区域 Oracle VM S11 区域 Solaris 10 实时迁移 V2V P2V Most of our customers migrate from old technologies to our latest servers to consolidate their applications and realign their workloads and resources with the business requirements. Migration is a core component of this process. How do upgrade and still run your existing applications without losing their conditions and services or do this securely over the network. Oracle Solaris Zones continue to be the best environment for deploying applications and consolidating services. Significant work has been done in Solaris 11 to better integrate this technology in the OS to allow organizations to easily meet changing requirements by quickly provisioning new Oracle Solaris Zones, or moving them from system to system or disk to disk within the same system as capacity or configuration needs change. In this example, we start with a traditional deployment of Solaris 10 on a single system … and we can then migrate the physical Solaris 10 installation to a virtual Solaris 10 Zone. Another physical to virtual migration path we could take would be to move our physical Solaris 10 installation to a virtual Solaris 10 Zone, but this time on the Solaris 11 operating environment. We could also fairly easily move our virtual Solaris 10 Zone from our Solaris 10 platform to Solaris 11. In this case, we’ve called out Oracle VM explicitly, but as we know, all T-series systems actually sit on a thin Oracle VM implementation. And this is where we can use our Live Migration feature to migrate an active domain to another physical machine while maintaining application services to users. Live Migration is available on SPARC-T systems and on-Chip Crypto helps accelerate the live migration markedly. 在 SPARC T 系列系统上进行在线迁移 芯片加密加速器可为在线迁移提供安全、线性加速 的加密

整体系统管理 虚拟化 维护 统一打补丁 监测 发现和开通设备 Comprehensive hardware management. Sun had a product called Ops Center, which was really around hardware management - How you provision the hardware, how you understand the health of the hardware, how do you update firmware and patches. Extended to virtualization, how do you provision virtual environments, operating systems, and so on. We've significantly enhanced this program to include telemetry back to Oracle as well as integration with OVM on x86 as well as SPARC and many other features. The big change here with Ops Center is we've recently announced that everybody under a hardware service contract with Oracle Systems products gets this product for free. This is actually an extraordinarily rich product because you can look at other virtualized infrastructure -- and, understand, virtualization comes with Oracle Systems as a standard feature, so you get OVM and you get Solaris and you get Oracle Linux on the x86 side, so we are giving you virtualization included with the systems with no license restrictions on the number of images you run. Then, for everyone under a service contract we're giving you all the management tools for virtualization, including discovery, provisioning, deployment, health monitoring, patching, and so on. And so this is an incredibly rich stack. If you think about being able to buy across Oracle’s server and storage portfolio, and also getting all the baseline management tools included.   发现和开通设备

Enterprise Mgr Ops Center Oracle 系统管理 — 揭开神秘面纱 集成设计，简化数据中心运营 服务器硬件（SPARC 和 X86） ILOM ENTERPRISE MANAGER OPS CENTER ORACLE VM 虚拟机管理程序 SOLARIS ORACLE 中间件、数据库和管理软件 全面掌控数据中心 Enterprise Mgr Ops Center  Oracle Solaris 11 Oracle VM Oracle 服务器 (ILOM) So if you take the lower stack into considerations al of components play a role in Data Center Management and they are well integrated of each other. Not only the tools read from the other components but also they can control the behavior of the others. Among all, the Enterprise Manager carries the larger weight with a rich set of features and capabilities not only for System Management but also for Cloud setup, deployment, management. Key Capabilities:   Uniform virtualization management across Oracle VM for SPARC and x86, as well as Solaris Zones Unified management for Solaris architecture including Solaris 11, Solaris Network virtualization and OS analytics Comprehensive management for SPARC T4 systems and Exalogic and Exadata engineered systems Extends Oracle Enterprise Manager 12c cloud lifecycle management with deep Oracle hardware integrations

Oracle 的优化解决方案 简化和加快应用部署 整个体系全面优化 由组织统一推动集成 硬件/软件同步发布计划 早期开发阶段测试 May 18, 2010 Oracle 的优化解决方案 简化和加快应用部署 整个体系全面优化 早期开发阶段测试 端到端的功能验证 互操作性测试 故障注入测试 负载/压力测试 性能和可扩展性测试 真实环境负载测试 选型与配置优化 由组织统一推动集成 硬件/软件同步发布计划 Copyright Oracle Corporation

加速 I’d like to close by providing a few examples of customers we’re working with.

单台服务器中速度最快的数据库服务器 Oracle 优势 带来闪存存储革命的服务器 10366254 8552523 tpmC IBM Power 780 3 节点集群 — AIX Oracle T5-8 Solaris SPARC T5-8 8552523 tpmC，0.55 美元/tpmC，2013 年 9 月 25 日发布。IBM IBM P780 Cluster 10366254 tpmC，1.38 美元/tpmC，2010 年 10 月 13 日发布。 参见基准测试披露幻灯片

7 倍 性价比突出的数据库服务器 Oracle 优势 190 万美元 27 万美元 带来闪存存储革命的服务器 数据库 每台服务器成本 IBM Power 780 3 节点集群 — AIX Oracle T5-8 Solaris 190 万美元 27 万美元 每台服务器成本

12 倍 速度最快的 Java 中间件服务器 Oracle 优势 99 万美元 27 万美元 更快、更高效的应用基础架构 16646 EjOPs 57422 12 倍 Oracle 优势 IBM P780 AIX Oracle T5-8 Solaris 99 万美元 27 万美元 成本 SPARC T5-8，57422.17 SPECjEnterprise2010 EjOPS；IBM Power 780，16646.34 SPECjEnterprise2010 EjOPS。 参见基准测试披露幻灯片

成果 I’d like to close by providing a few examples of customers we’re working with.

数据集成服务提供商 Oracle 数据库负载的终极整合 成果 公司 机遇 每个处理器的性能比 IBM Power 750 高 3.4 倍 数据集成服务器提供商 — 北美 通过整合 IBM 和 HP 系统实现更高性能的 中端 SPARC T5-8 机遇 证明可将 IBM Power 系统极限整合到一台 SPARC T5-8 上 Informatica、Oracle Business Intelligence、 Oracle 融合中间件和 Oracle Database 11gR2 的 Oracle 软硬件一体化解决方案 成果 每个处理器的性能比 IBM Power 750 高 3.4 倍 Oracle DB 11gR2 数据加载速度快 1.2 倍 使用 T5-8 可将整合率提升至 6:1 Another example of our advantage over IBM and HP. One of Oracle’s customers in North America, a services provider and data integration software application developer tested SPARC T5-8. The results show a 3.4x performance running Oracle database queries on Oracle 11gR2 and Solaris 11. The consolidation ratio of IBM POWER 750 and HP x86 systems running Linux was 6X. Pricing: Including the AIX licensing and LPAR (Power VM), price per server is $110K, or $441K for the four P750s T5-8 $268K Price difference is 1.8x advantage T5-8 3X Advantage based on performance per processor: 3.4x better performance per processor Baseline improvement in query performance (without using parallel query) on SPARC T5 ranges from 34% to 70% against similarly configured (virtualized environment using same number of cores) Power 7 and in range of 2x compared to older Power 5 systems. The comparison was to P7 system (and not to P5 system) Let us get specific about the system for our own info. The message here is T5 had a 35% to 70% better performance per core over P7. 1.7 *2 = 3.4 Customer was able to successfully migrate and test multiple databases and applications from other UNIX (AIX and Linux) system to a  single Oracle SPARC T5-8 server configured with Solaris 11 and Oracle VM for SPARC. 1.2X Faster Data load: Customer migrated 7+TB of database from AIX to Solaris/SPARC T5 under two hours, meeting their SLA of not more than 3 hours of downtime. 6:1 Database consolidation: Qty 4 x 750s + 2 x HP x86s = 6 6:1 databases into a single T5-8

数据集成服务提供商 Oracle 数据库负载的终极整合 IBM Power 750 HP x86 SPARC T5-8 Oracle Solaris 11 占用空间减少至原来的 1/6 Oracle 一体化解决方案 出色的可扩展性 和虚拟化

电信提供商客户 T5 大幅提高用户响应速度 成果 公司 响应速度提高 2 倍 批量处理速度提高 2 倍 占用空间减少至原来的 1/8 机遇 欧洲大型电信提供商 整合原有 SPARC 系统，需要最高水平的可 用性和可靠性 机遇 展现 Finnova 财务应用在 SPARC 上的性能 在 SPARC T5-4 中端服务器上测试 Oracle 优化的软件体系的效果 成果 响应速度提高 2 倍 批量处理速度提高 2 倍 占用空间减少至原来的 1/8 Baseline for performance test: M8000 15X Advantage: 7x based on list price comparison between M8000 and T5-4 M8000 list price justification 4so 2.88GHz $1M T5-4 list price justification 4so 1TB $149K 2.5X Application Throughput: Highest test result comparison, application server response time: T5-4 = .09 and M8000 = .25, therefore 2.7X increase Additional results: T5-4 = .79 and M8000 = 1.82 therefore 2.3X increase T5-4 = .66 and M8000 = 1.29 therefore 1.95X increase Total T5-4 is 2.18 total for M8000 = 3.36 therefore 2.2x increase 2x faster batch: Throughput increase based on comparison of month end numbers 8X Space reduction: Done by RU 42 vs 5 RU

Oracle Solaris/SPARC 卓越的可扩展性和虚拟化功能 电信提供商客户 T5 大幅提高用户响应速度 Oracle Solaris/SPARC 卓越的可扩展性和虚拟化功能   SPARC T5-4 Oracle Solaris 11 SPARC M8000 出色的可扩展性 和虚拟化 重新定义数据中心中端服务器

最全面的行业解决方案 Just before 航空与国防 前 11 家 汽车 300 多家 OEM 和供应商 化工 全球前 10 家 中的 5 家 通信 前 20 大 服务提供商 消费品 前 20 家 教育与科研 前 10 大院校 中的 9 所 工程与建设 财富 500 强前 5 家 中的 4 家 金融服务 全球前 10 大 银行 医疗保健 300 多家 龙头提供商 高科技 前 25 大 电子 OEM 工业制造 全球前 10 家 中的 9 家 保险 全球保险公司 前 50 家中的 49 家 生命科学 前 25 家制药公司 中的 20 家 媒体/娱乐 财富全球 500 强 石油与天然气 前 7 大企业 中的 6 家 Just before Here’s a quick look at some of the industries we serve. This scorecard is a testament to Oracle’s solid presence across a broad range of industries as well as customer successes we have helped build. 专业服务 全球前 10 大 IT 服务公司 中的 9 家 旅游与运输 前 5 家航空公司 中的 3 家 零售 前 20 家 公用事业 前 20 家 公共部门 1500 多个 组织机构

ORACLE 体系与 SPARC / Oracle Solaris 数据中心现代化 Oracle 运行在Oracle上 Oracle工程一体化 软件嵌入芯片 云和虚拟化 优化 简化部署 更易于管理 更敏捷 最低的 TCO 简化 速度最快的单服务器 TPCC SPECjEnterprise2010 现已创造 9 项世界纪录 加速

oracle.com/sparc oracle.com/sparcTCO

Graphic Section Divider

基准测试真实性证据 如需了解所有世界纪录，请访问 www.oracle.com/benchmarks 和 blogs.oracle.com/bestperf/ 信息来源：Transaction Processing Performance Council 网站 www.spec.org，截至 2013 年 3 月 26 日。TPC Benchmark C、tpmC 和 TPC-C 是 Transaction Performance Processing Council (TPC) 的商标。在 SPARC T5-8 (8/128/1024) 上运行包含分区特性的 Oracle Database 11g 第 2 版企业版，8552523 tpmC，0.55 美元/tpmC，2013 年 9 月 25 日发布。 SPEC 和基准测试名称 SPECjEnterprise 是 Standard Performance Evaluation Corporation 的注册商标。比较结果来自 www.spec.org，截至 2013 年 3 月 26 日。SPARC T5-8，57422.17 SPECjEnterprise2010 EjOPS；IBM Power 780，16646.34 SPECjEnterprise2010 EjOPS。 SPARC T5-8 服务器配置：SPARC T5-8 服务器基础包，8 个 SPARC T5 16 核处理器，128 个 16GB-1066 DIMM，2 个 600GB 10K RPM 2.5 英寸 SAS-2 HDD，4 根电源电缆，报价为 268742 美元。IBM Power 780 配置：IBM Power 780:9179 Model MHB，4 个 3.8GHz 16 核处理器， 64 个单处理器激活，4 个 CEC 机柜（包括 IBM 机箱、I/O 背板和系统中间背板），16 条 0/32GB DDR3 内存 (4x8GB) DIMMS-1066MHz Power7 CoD 内存，12 个 1 GB DDR3 Power7 内存激活，5 个 100GB DDR3 Power7 内存激活，1 个磁盘/介质背板。2 个 146.8GB SAS 15K RPM 2.5 英寸 HDD（仅适合 AIX/Linux），4 个 AC 电源 (1725W))，定价 992023 美元。信息来源：Oracle 价目表和 IBM.com，收集日期： 2013 年 3 月 25 日。性价比 = 12.7 = (992023/16646.34)/(268742/57422.17)。 信息来源：Transaction Processing Performance Council 网站 www.tpc.org，截至 2013 年 3 月 26 日。在 SPARC T5-8 (8/128/1024) 上运行包 含分区特性的 Oracle Database 11g 第 2 版企业版，8552523 tpmC，0.55 美元/tpmC，截至 2013 年 9 月 25 日。3 节点 IBM Power 780 9179- MHB 服务器，Power 7 3.86 GHz (8/64/256) 运行 IBM DB2 9.7，10366254 tpmC，1.38 美元/tpmC，截至 10 年 10 月 13 日。性价比 = 2.5 = (1.38/0.55)。 SPARC T5-8：2GB - 268742 美元；8 个 3.6 GHz T5 CPU，2024 GB 内存。IBM Power 780：1908530 美元，8 个 3.92 GHz 64 核 POWER7 CPU，2024 GB 内存；7 倍的价格优势 = (1908530 美元/268742 美元) 价格来自 IBM.com 和 Oracle 价目表

基准测试真实性证据 SPARC M5-32 配置：SPARC M5-32 服务器基础包，32 个 M5 6 核处理器，64 个 16GB-1066 DIMM，6 个 600GB RPM 2.5 英寸 SAS-2 HDD， 6 根电源电缆，报价为 1045962 美元。IBM Power 795 配置：包含 CEC 机柜和 BPA 的 IBM Power 795 主系统机架、4 个 4.0GHz 32 核处理器， 2 个适用于 Power 7 的 64 处理器内核激活，2 个 1024 (8x(4x32GB)) DDR3 DIMM-1066MHz，48 个 1GB DDR3 Power7 内存激活，20 个 100GB DDR3 Power7 内存激活，1 个 12X I/O 26 SFF 驱动器架，2 个 600GB SAS 10K RPM HDD，1 个 DVD/磁带 SAS 外部存储单元，1 个 机架安装式硬件管理控制台，1 个电源线缆组：BPD1 到第一个处理器节点，1 个电源线缆组：BPD1 到第二个处理器节点，1 个电源线缆组： BPD1 到第三个处理器节点，1 个电源线缆组：BPD1 到第四个处理器节点，6 个 Bulk Power Regulator，2 个 Bulk Power Distribution Assembly）。报价为 2719336 美元，信息来源：ibm.com，收集日期：2013 年 3 月 25 日。 为了了解能节省的成本，我们进行了一项为期 5 年的 TCO 分析：用针对 Oracle 数据库的 Oracle 优化解决方案（使用两台 Oracle SPARC T5-4 和存储）替代 IBM 的十个 (10) IBM Power P570、Power 6+ 处理器和老式 SAN 存储。该分析的前提是替换现有部署，因此需要 Oracle 数据库 企业版和 Oracle Real Application Cluster 许可。分析结果证明，与原有 IBM 系统相比，针对 Oracle 数据库的 Oracle 优化解决方案存在巨大优 势。5 年内节省总拥有成本 480 万美元，5 台服务器整合为 1 台。电力和散热成本降低至原来的 3.5 分之一。 SPEC 和基准测试名称 SPECjEnterprise 是 Standard Performance Evaluation Corporation 的注册商标。比较结果来自 www.spec.org，截至 2013 年 3 月 26 日。SPARC T5-8，57422.17 SPECjEnterprise2010 EjOPS；IBM Power 780，16646.34 SPECjEnterprise2010 EjOPS。 SPARC T5-8 服务器配置：SPARC T5-8 服务器基础包，8 个 SPARC T5 16 核处理器，128 个 16GB-1066 DIMM，2 个 600GB 10K RPM 2.5 英寸 SAS-2 HDD，4 根电源电缆，报价为 268742 美元。IBM Power 780 配置：IBM Power 780:9179 Model MHB，4 个 3.8GHz 16 核，64 个 单处理器激活，4 个 CEC 机柜（包括 IBM 机箱、I/O 背板和系统中间背板），16 条 0/32GB DDR3 内存 (4x8GB) DIMM-1066MHz Power7 CoD 内存，12 个 1 GB DDR3 Power7 内存激活，5 个 100GB DDR3 Power7 内存激活，1 个磁盘/介质背板。2 个 146.8GB SAS 15K RPM 2.5 英寸 HDD（仅适合 AIX/Linux），4 个 AC 电源 (1725W))，定价 992023 美元。信息来源：Oracle 价目表和 IBM.com，收集日期：2013 年 3 月 25 日。