Everyone is talking about cloud. Rarely have I seen such a high level of interest, expectation, excitement about a technology topic. It’s even crossed over from technology publications to the mainstream, as you can see here in BusinessWeek, The Economist and also the Wall Street Journal. When that happens, it means enough people have interest in it.
This is Gartner‘s “Hype Curve” for cloud computing. Cloud computing is the most hyped subject in IT today. Note what is right up there at the “Peak of Inflated Expectations” – Cloud Computing. And it’s about to roll into the “Trough of Disillusionment” But also note what technologies are on the “Slope of Enlightenment.” These are technologies that have been around a while. Proven, mature, real, widely adopted. Grid Computing. Utility Computing. Virtualization. SaaS. Cloud is related to and based on these. Cloud is the evolution and convergence of these.
One of the areas of confusion is the definition of Cloud Computing. There are many definitions of Cloud Computing out there. Here is one of them that seems to represent the most commonly held view. It’s from the National Institute of Standards (NIST) and seems to be gaining in popularity, not only in the US, but also the rest of the world as well. The definition is essentially about “on-demand access to a shared pool of computing resources.” Breaking it down, cloud computing is composed of: 5 essential characteristics 3 service models 4 deployment models The 5 essential characteristics are key: On demand self-service – provisioning, monitoring, management control Resource pooling – implies sharing and a level of abstraction between consumers and services Rapid elasticity – the ability to quickly scale up/down as needed Measured service – metering utilization for either internal chargeback (private cloud) or external billing (public cloud) Broad network access – typically means access through a browser on any networked device I’ll cover the 3 service models and 4 deployment models on the next few slides.
“ Software as a Service” generally refers to applications that are delivered to END-USERS over the Internet. There are hundreds of SaaS providers out there covering a wide variety of applications. Oracle CRM On Demand is an example of a SaaS service. Another example is Salesforce.com “ Platform as a Service” generally refers to an application development and deployment platform delivered as a service to DEVELOPERS, allowing them to quickly build and deploy a SaaS application to end-users. These platform are often built on a grid computing architecture and include database and middleware. They are often specific to a language or API. For example Google AppEngine is Java and Python. EngineYard is Ruby on Rails. Salesforce.com’s Force.com is a proprietary variation of Java. Finally, “Infrastructure as a Service” generally refers to computing hardware (servers, storage and network) delivered as a service. This typically includes the associated software as well: operating systems, virtualization, clustering, etc. The best known example of this is Amazon Web Services, which offers Elastic Compute Cloud (EC2) for compute servers and Simple Storage Service (S3) for storage.
Animated slide. Let’s look closer at the distinction between public and private clouds. [CLICK] A public cloud is shared by multiple tenants, whereas a private cloud is for the exclusive use of a single organization. A public cloud is hosted and managed by the cloud service provider, and a private cloud is controlled and managed by in-house IT (of course, it’s also possible to outsource this, so there are such things as “hosted private clouds” or “virtual private clouds” but for the sake of simplicity, it’s easier to think of private clouds as in-house. A third observation is that public clouds usually offer a very limited variety of offerings, in order to be efficient, while a private cloud may need to provide a large number applications. Within a large enterprise, there are typically hundreds to thousands of apps. The NIST model includes “community clouds” which are essentially semi-private clouds for use by a group related organizations, such as all the schools in the University of California system, all the branches of the military, or all the parts suppliers to Ford or GM. And a hybrid cloud is some combination of the other three…typically for a single application. (if an organization has 1 app in a private cloud and a different app in a public cloud, that’s not considered a hybrid cloud). [CLICK] Each has its own unique advantages, and they have some common advantages as well. Because both public and private clouds are based on virtualization and grid computing, they enjoy high efficiency and utilization rates, elastic capacity for limitless scale-out and pay-as-you-go equipment procurement, and also high availability for maintaining high user service levels and business continuity. Public clouds are often faster and cheaper to get started, since there’s nothing to install. They offer economies of scale which the provider can pass on to customers. They don’t require IT to manage and administer, update, patch, etc. And they are paid for as Operating Expense, which can be simpler from a budgeting standpoint. Private clouds offer greater control over security and data privacy, compliance (this can be a big issue since there are some regulatory requirements about where data resides, audit trails, etc. that public clouds cannot meet today), and also quality of service, since private clouds can manage network bandwidth and implement optimizations that public clouds don’t allow. Private clouds also provide easier integration with other systems that are on-premise. They are potentially lower cost over the long term…breakeven is in 2 or 3 years. After that, public clouds become more expensive. And private clouds are paid for as both Capital Expense (with depreciation) and Operating Expense. Enterprises will make these trade-offs and will likely run a mix of public and private clouds. Even Oracle, which operates one of the biggest private clouds internally, also uses Amazon EC2 for some things, such as marketing demos. One popular use case is Dev & Test…engineering can use public cloud resources to set up development and test machines without waiting for IT to set them up. Another interesting use case is doing disaster recovery offsite in a public cloud.
In this Forrester survey, 44% of large enterprises (more than 1000 employees) are interested in building an internal (private) cloud. The number is smaller for medium sized businesses and even smaller for small businesses. Very intuitive, since it’s the larger companies with sufficient scale and IT skills to build a private cloud.
What are the key benefits that enterprises see in Cloud Computing? Here’s are some recent results from a survey by IDC. Benefits : the top reason to use cloud computing is speed/ease of deployment, and the next 3 are all related to lower costs. Issues : Security is the top issue. The next 2 (Perf & Avail) relate to Quality of Service. The next 2 relate to concerns about how well the cloud application fits the business requirements. There is also concern about long-term costs, lock-in and regulatory compliance.
Real, new capabilities: On demand self-service, elastic scalability/capacity, measured service to enable pay-per-use Established technologies: grid, virtualization, dynamic provisioning “ Cloud is evolutionary, not revolutionary” (grid computing, virtualization, SOA, shared services, SaaS, outsourcing, broadband networks, browser as the platform) Benefits: speed/agility and cost Concerns: security, compliance, QoS, integration, lock-in, long-term costs Enterprises will use a mix of public and private clouds. Enterprise IT departments will build private clouds to provide the benefits of public clouds (speed, cost) while mitigating the serious concerns (security, compliance, etc.). Some apps will run on public, while others will run on private, while others will stay non-cloud. Adoption will be gradual, taking place over several years. So given this perspective, what is Oracle’s Cloud Computing Strategy?
This slide has animations/builds. So, what is Oracle’s overall cloud computing strategy? Our objectives are first to ensure that cloud computing is fully enterprise grade, that is, that is high performance, scalability, reliability, availability, security and standards-based for portability and interoperability. All the “ilities”… Secondly, we will support both public and private cloud computing in order to give customers choice. To accomplish this objectives, we have a 2-pronged strategy: [CLICK] First, we offer Technology to enable enterprises to build private clouds or run in public clouds [CLICK] Second, we offer Apps that are deployed in either a private shared services environment or public SaaS Let’s drill down into each of these…
This slide has animations/builds. First for Technology: [CLICK] We offer Technology to enterprises building Private Clouds. We will talk in depth about the “Oracle Private PaaS” [CLICK] We also offer enterprises the ability to run Oracle Technology in public clouds like Amazon and Rackspace. Then for Applications: [CLICK] We offer Oracle Applications running in a shared services private cloud, and [CLICK] We also offer a number of Applications as public cloud services from our Oracle On Demand business. We’re now going to talk more about each of these 4 areas. The axes are Public and Private, Tech and Apps. But we will focus more on the Technology side in this presentation and the following presentations.
First I’ll talk about our strategy for Technology for Private Clouds , which is our offering: Oracle Private PaaS
Oracle Private PaaS, the WHAT, the WHY and the HOW
Let’s start with the WHY Enterprise IT departments will build Private PaaS because they get the benefits of cloud computing (agility/speed, efficiency/cost) without the risks of public clouds (security, compliance, QoS), and because a platform approach encourages component re-use and standard shared services, which makes it much faster, easier and cheaper for developers to build apps by assembling components instead of writing everything from scratch. A platform offers the best combination of flexibility and control. An IaaS provides only a “flashing OS cursor.” It doesn’t do anything until the user builds and deploys something on top of that operating system. It’s very flexible, but it’s a lot more work than a platform approach.
Now let’s talk about WHAT is a Private PaaS. A Private PaaS is made up of a number of critical building blocks. Oracle has the most comprehensive set of building blocks in the industry, the most “complete, open and integrated” set of building blocks. From the bottom up, this includes Oracle VM for server virtualization, Oracle Enterprise Linux our OS, the Oracle Database grid (made up of RAC, ASM, In-Memory Database Cache, and other database options and features). Then on top of that, Oracle offers our application grid, which includes WebLogic Server, Coherence, Tuxedo and JRockit, and on top of that, a number of value-added services: SOA and BPM for integration and process management, identity and access management for security, and WebCenter our portal for user interaction. We also offer very comprehensive “Cloud Management” capabilities based on Oracle Enterprise Manager. EM has very comprehensive capabilities to manage the full “Cloud Platform” stack including middleware, database, OS and virtualization. For example, Real User Experience Insight (RUEI) enables us to manage top-down from the application end-user’s perspective things like performance, availability and behavior patterns…something that’s useful for SLA/QoS management for private clouds. Our second Keynote explains Private PaaS in more depth, and we have a separate session to talk more about Cloud Management.
I would like to use this slide showing the lifecycle of how a private cloud would work within an enterprise. Note the different roles. 1. You have [click] a central IT function that initially sets up the platform, [click] installing the private PaaS, [click] building the self-service Web-based interface perhaps as a portal, and [click] creating some shareable components (may be services, processes or UI components). 2. Next, a departmental app owner [click] can take advantage of the PaaS and shared component to more quickly assemble the app [click] and deploy it through self-service [click]. If their role entitles them to make that request, it is automatically provisioned. If not, it gets routed to their management and/or IT for workflow approval…just like a procurement process. 3. Then you have [click] the users of the department apps, who may also employees within that department, other departments, partners, or customers. 4. Finally, there’s application administration [click]. An app owner goes through self-service to manage the app in the private cloud. The owner can monitor and manage the app, adjust capacity if necessary and tracks usage (metering) and how much they are being charged for use of the cloud resources. So, this PaaS shows some of the key characteristics of cloud computing: self-service, shared services, dynamic provisioning, elastic scalabilty and metering/chargeback.
This slide has animations/builds. So we’ve covered the WHAT and the WHY of Enterprise Private PaaS. Now let’s talk a bit about HOW. We believe that enterprises will want to EVOLVE their current IT infrastructure to become more “cloud-like” – to become a better internal service provider to the lines of business, BUs, departments – to provide greater agility and responsiveness to business needs, higher quality of service in terms of latency & availability, and lower costs and higher utilization. This evolution will take time. Not only is the available technology evolving and advancing, but enterprises are also working on the new policies and processes needed. In many cases, the technical building blocks for cloud computing are available in advance of enterprise readiness, so we think that enterprises will evolve towards the right at different rates. The first step that many enterprises are taking is to move from a “Silo’ed” environment to a “Grid” or virtualized environment –moving from a dedicated, rigid, physical structure for each application to a virtual environment with shared services, dynamic provisioning and standardized configurations or appliances. This trend is very strong right now. Many enterprises are leveraging Grid and virtualization technologies to consolidate and reduce costs. Oracle has a very strong and complete offering for Grid, with products in the database and middleware layers, such as RAC, TimesTen, WebLogic and Coherence, plus Oracle VM for server virtualization and Enterprise Manager for managing the entire stack. [CLICK] From here, enterprises can evolve to a self-service and pay-per-use environment, similar to how Amazon works. A user goes to the employee portal, signs in, makes a request for a virtual machine(s) with a certain amount of CPU, memory and disk, picks a VM image for database or middleware, then clicks “submit.” If that employee’s role and entitlements allow him to have that amount of IT resource, then it auto-magically gets provisioned without an IT person being involved. If not, perhaps his request gets routed to his manager and/or IT for workflow approval. In 10 minutes, they are up and running with a full “private PaaS.” After he deploys the app, the system has policy-based resource management to automatically make capacity adjustments, and the employee’s business unit gets an internal charge every month based on how much IT resources they consumed. To make all that happen, the enterprise must have policies and processes defined, and the technology must be able to support it. [CLICK] Meanwhile, public clouds are also evolving. There are already many different public cloud offerings. They are at all the layers: SaaS, PaaS, IaaS. They are specialized and isolated. CLICK] Ultimately, we think the evolution will move to “hybrid clouds” where a single application can span both private and public clouds and is managed in a federated manner “through a single pane of glass.” For this to happen, there need to be standards for interoperability and portability, and there needs to be technology to support such interoperability. The notion of “cloud bursting” is very compelling to CIOs. This is when the steady-state workload runs on the in-house private cloud, on hardware and data centers owned by the enterprise. But when there’s a peak in the workload, it can dynamically burst out to a public cloud and take advantage of that capacity. When the peak is over, return that capacity back to the pool and shed that cost. We think cloud bursting is still a ways off in the future, but it certainly is very compelling.
Grid Computing is the first step in the evolution to cloud. Grid is about virtualized and sharing resources, and distributing and redistributing capacity as needed. Oracle has applied Grid principles to all levels of the modern software stack: storage, database and middleware. Traditional Computing Infrastructure: Silo’d, underutilized, inefficient to mange Grid infrastructure: standardized, consolidated, virtualized, highly automated with rightsized capacity
Virtualization and Clustering are the two key technologies that make up Grid Computing. These technologies are complementary. Unlike what some other vendors would have you believe, Cloud Computing is NOT just server virtualization. Clustering is also an important enabler to cloud computing. Virtualization: makes a single computer look like many computers. The size and power are variable/configurable. Virtual machines can be migrated without downtime Virtual machines enable far more detailed accounting of which applications, LOBs, customers are using IT resources. Clustering with RAC or WebLogic is also a type of virtualization. It makes many computers (or even virtual machines) look like a single resource. Huge databases and middleware tiers can be built using powerful, low-cost, high volume components (like blades or rack servers) Redundancy of clusters enables high-performance and scalability through parallel operations Redundancy also enables inherent high availability, as clusters can survive one or more node failures
One of the “essential characteristics” (as defined by NIST) of Cloud Computing is “resource pooling” or sharing. Grid is what enables resource sharing and consolidation. Many enterprises (I would say more than half) are doing this today: using virtualization and grid computing to share and consolidate, to save money. A grid approach enables consolidation of IT resources at many levels. Consolidation to clusters or virtual machines means a better use of IT resources. Take advantage of complementary workloads peaks - clustered or virtualized resources can be easily re-configured to meet peak load requirements of any particular application. Higher utilization rates and efficiency – workload management means less hardware can service more applications. Lower CapEx and OpEx – fewer computers, few to manage and cool Green footprint – fewer computers, less power used, less cooling required. Also, high volume components are getting more and more energy efficient, so server refresh improves green advantages.
Another “essential characteristic” of Cloud Computing (according to the NIST definition) is “Elastic Scalability.” Grid enables elastic scalability through the ability to add and remove computing resources with no interruption in service. Oracle provides clustering and virtualization at all levels: DB, MW, Storage Pay as you go: predictable scale out Easier to capacity plan, automation makes adding additional capacity faster and easier. Buy only as much capacity as you need this year, not for the next 3 years.
High Quality of Service isn’t a NIST “essential characteristic” but I think it’s expected for any mission-critical application. Grid also enables high QoS: predictable performance at any scale and high availability. Grid employs redundancy to survive one or more component failures and ensure high availability and predictable performance at any scale.
Oracle has the most comprehensive stack of Grid Computing technologies in the industry. Companies who what to evolve toward Cloud Computing need to start with a solid foundation based on Grid Computing. From this, they can add self-service, chargeback, automated capacity adjustments in order to evolve to a Private Cloud. We’ll talk a lot more about what a Private Cloud is in Keynote #2 coming up right after this. That keynote is called “Essential Building Blocks for Private PaaS.”
To illustrate how enterprise are evolving to cloud computing, I would like to tell you a bit about Oracle’s own IT department. We ourselves are a large enterprise (85k employees, 4 million external users, 42,000 servers, 10,000 TB of storage). We have 4 different parts of IT. We have a part that operates the business applications we use to run our business (ERP, CRM, etc.)...this is similar to what every enterprise would have. The other 3 pieces may be different from yours. We have Oracle University, which requires us to setup systems to train thousands of people every week. We have Oracle On Demand, where we offer SaaS apps and managing/hosting services. And finally, we have a very large Development group, which requires a lot of computer resources for dev/test.
Of these 4 pieces of Oracle IT, the Development area is furthest along in the evolution to cloud. They have a full “self-service private cloud,” which provides tremendous value and competitive advantage for Dev/Test. A developer from our Database, Middleware or Enterprise Manager teams can submit their jobs through a self-service application. That job gets automatically queued, assigned to a host or set of hosts, where it runs. When done, results are sent back to the Developer. The Developer has no idea where their job is running, but they just know it will run. The system is highly automated, requiring very little IT administrative support, and fully self-service for the developer, making it very fast and efficient for them.
This data applies to our Intel x86/x64 Farm. This Farm is used by engineering for the development of Database, Fusion Middleware, and Enterprise Manager products on Linux (OEL, RHAT,Suse) 32 and 64, Windows (XP, 7, Server 2003, Server 2008, Vista) 32 and 64, and Solaris on Intel 64. We also have a variety of configurations (e.g. 2-node clusters, 4-node clusters, 8GB of Memory systems, 4GB of Memory systems, etc.). We have ~6,000 virtual servers across ~2,600 physical servers. The Farm is used by over ~3,500 developers. THERE ARE ONLY 5 ADMINS FOR THE ENTIRE FARM. For the Linux OEL4 32, which is the biggest portion of the Farm being LINUX our base dev environment, we are able to process over 70K hours per day. This translates into over 45,000 jobs processed. &quot;jobs&quot; are either execution of product builds or product tests. These test run for an average of close to ~2 hrs. The utilization rate of these serves is about 80% over 7 days a week. Monday through Friday by the way is almost 100% as people take a break over the weekend :) . Normally, Dev/Test machines are among the lowest in utilization, usually about 5-10%, but we run at over 80%. The benefits have been the following: - A significant boost in development productivity. A developers could simply not do what he's able to do now without the Farm. It is self-service system when developers simply need to &quot;push a button&quot; to submit their jobs. - Cleaner code lines as code is thoroughly tested before being committed to the product. - Speed of development has been increased. This whole system provides us (i.e. Oracle Development) a significant competitive advantage over other software shops. That's why I hesitant, by the way, to talk much about it :) We also have similar Farms for HP-UX (Itanium and PA-RISC), AIX, Solaris Sparc, Intel Itanium, zLinux, LoP.
Oracle University is at a different point in the evolution to cloud. They are at the “Grid” stage, so not quite as far along as Development’s self-service private cloud. This situation is fairly unique. They have to stand up over 2300 VMs every week for students to use, and then they tear them back down again at the end of the week. So this is an extremely dynamic environment, much more so that yours probably. By leveraging Oracle VM and grid computing, they were able to save 90% of the hardware, which of course saves power, floor space, and human admin time. These are some impressive metrics.
We just spent a lot of time talking about Enterprise Private Clouds, about the what, why and how. Now we are going to cover the other 3 parts of our strategy, but we’ll do it quickly for now. We will come back and touch on these more in other parts of our Cloud Computing Forum event today.
At Oracle OpenWorld 2008, Oracle announced that we would enable customers to run Oracle products in public clouds. The first was Amazon’s Elastic Compute Cloud (EC2), part of Amazon Web Services (AWS), and our intent is to support other public IaaS cloud providers as well. Specifically, we allow customers to use their existing licenses or purchase new licenses for Oracle Database, Oracle Fusion Middleware and Oracle Enterprise Manager, and run those in either the own data centers or in EC2. Oracle will provide support for our products running in EC2. You will soon be able to run Oracle Technology in Rackspace as well. They are building a public PaaS offering based on Oracle’s Technology stack including WebLogic Server, Oracle Database RAC, Oracle Enterprise Linux and Oracle VM.
Now let’s switch and talk about cloud strategy for Applications. In the Private case, we’re talking about the broad portfolio of Oracle applications, both horizontal and industry-specific, running on a Shared Services Private PaaS.
Oracle On Demand is the name of a wide spectrum of service offerings from Oracle that give customers the choice of deploying on-premise in the own data centers or in Oracle’s data centers, managed by Oracle or by the customer, and how maintenance and optimization get done. At the far right is the traditional software license and on-site deployment model where the customer licenses the software and deploys and manages it themselves in their own data centers. Moving over towards the left, Oracle On Demand offers a remote management service where we can manage the software for you – the software is still deployed in your own data centers. Oracle will also do hosting and management at Oracle’s data centers – in this model, the customer buys a perpetual license and annual maintenance, and pays Oracle a fee to provide hosting/management services. The two models at the far left of this are both considered SaaS, meaning the customer is paying-per-use of the software. A rental. Or a longer-term lease. One is a single-tenant model where the customer gets a dedicated system, which can be optimized for the customer and for which the customer has a degree of control, such as specifying when maintenance gets done for example. In a multi-tenant model, the customer shares resources with other customers, so the cost is lower, but the vendor must treat the group of customers exactly the same in order to get the cost efficiencies out of this model. Oracle On Demand offers the 4 deployment options towards the left side of this picture.
To conclude this session, I’d just like to summarize with a few final points: First, Oracle provides most complete, open and integrated cloud vision, strategy and offerings in the industry. We offer a very comprehensive set of building blocks for building and managing public and private clouds from applications to disk, as you will see in the next Keynote. Second, Cloud is the evolution of capabilities Oracle has been working on for more than a decade: grid computing, virtualization, shared services and management systems. Oracle helps enterprise IT departments evolve to become private cloud service providers. Finally, Oracle’s cloud computing strategy is to offer: Technology to build private clouds or run in public clouds Applications deployed in private shared services environment or via public SaaS
Transcript of "Cloud Computing"
Cloud Computing: What, Why and How Michał Kuratczyk principal solution architect
The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remain at the sole discretion of Oracle.
Cloud Is at the Peak of the Hype Curve Source: Gartner "Hype Cycle for Cloud Computing, 2009" Research Note G00168780
NIST Definition of Cloud Computing <ul><li>Cloud computing is a model for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. </li></ul><ul><li>This cloud model promotes availability and is composed of: </li></ul>Source: NIST Definition of Cloud Computing v15 <ul><li>3 Service Models </li></ul><ul><li>SaaS </li></ul><ul><li>PaaS </li></ul><ul><li>IaaS </li></ul><ul><li>4 Deployment Models </li></ul><ul><li>Public Cloud </li></ul><ul><li>Private Cloud </li></ul><ul><li>Community Cloud </li></ul><ul><li>Hybrid Cloud </li></ul><ul><li>5 Essential Characteristics </li></ul><ul><li>On-demand self-service </li></ul><ul><li>Resource pooling </li></ul><ul><li>Rapid elasticity </li></ul><ul><li>Measured service </li></ul><ul><li>Broad network access </li></ul>
SaaS, PaaS and IaaS <ul><li>Applications delivered as a service to end-users over the Internet </li></ul>Infrastructure as a Service Platform as a Service Software as a Service App development & deployment platform delivered as a service Server, storage and network hardware and associated software delivered as a service
Public Clouds and Private Clouds I N T E R N E T Public Clouds IaaS PaaS SaaS I N T R A N E T Private Cloud Users <ul><li>Public Clouds: </li></ul><ul><li>Lower upfront costs </li></ul><ul><li>Economies of scale </li></ul><ul><li>Simpler to manage </li></ul><ul><li>OpEx </li></ul><ul><li>Private Cloud: </li></ul><ul><li>Lower total costs </li></ul><ul><li>Greater control over security, compliance & quality of service </li></ul><ul><li>Easier integration </li></ul><ul><li>CapEx & OpEx </li></ul><ul><li>Both offer: </li></ul><ul><li>High efficiency </li></ul><ul><li>High availability </li></ul><ul><li>Elastic capacity </li></ul><ul><li>Used by multiple tenants on a shared basis </li></ul><ul><li>Hosted and managed by cloud service provider </li></ul><ul><li>Limited variety of offerings </li></ul><ul><li>Exclusively used by a single organization </li></ul><ul><li>Controlled and managed by in-house IT </li></ul><ul><li>Large number of applications </li></ul>IaaS PaaS SaaS
44% of Large Enterprises Are Interested In Building An Internal Cloud Source: Cloud Computing, Compute-As-A-Service: Interest And Adoption By Company Size, Forrester Research, Inc., February 27, 2009
Why Are Enterprises Interested in Cloud? What Are the Challenges Enterprises Face? Speed Cost QoS Fit Security Benefits Challenges/Issues Source: IDC eXchange, "IT Cloud Services User Survey, pt. 2: Top Benefits & Challenges," (http://blogs.idc.com/ie/?p=210), October 2, 2008
Cloud Computing: Oracle’s Perspective <ul><li>Characterized by real, new capabilities, but based on many established technologies </li></ul><ul><li>Compelling benefits as well as serious concerns </li></ul><ul><li>Enterprises will adopt a mix of public and private clouds </li></ul>
Oracle Cloud Computing Strategy Public Clouds IaaS PaaS SaaS I N T R A N E T Private Cloud Users <ul><li>Our objectives: </li></ul><ul><li>Ensure that cloud computing is fully enterprise grade </li></ul><ul><li>Support both public and private cloud computing – give customers choice </li></ul>IaaS PaaS SaaS I N T E R N E T Offer Technology to build private clouds or run in public clouds IaaS PaaS IaaS PaaS Offer Applications deployed in private shared services environment or via public SaaS SaaS SaaS
Oracle Cloud Computing Strategy Public Clouds IaaS PaaS SaaS I N T R A N E T Private Cloud Users IaaS PaaS SaaS I N T E R N E T IaaS PaaS IaaS PaaS SaaS SaaS Oracle Technology in public clouds Oracle Applications On Demand Oracle Applications Oracle Private PaaS
Oracle Cloud Computing Strategy Public Clouds IaaS PaaS SaaS I N T R A N E T Private Cloud Users IaaS PaaS SaaS I N T E R N E T IaaS PaaS IaaS PaaS SaaS SaaS Oracle Technology in public clouds Oracle Applications On Demand Oracle Applications Oracle Private PaaS
Why Enterprise Private PaaS <ul><li>Why Cloud? </li></ul><ul><ul><li>Agility and speed </li></ul></ul><ul><ul><li>Efficiency and cost </li></ul></ul><ul><li>Why Private? </li></ul><ul><ul><li>Security </li></ul></ul><ul><ul><li>Compliance </li></ul></ul><ul><ul><li>Control (particularly over QoS) </li></ul></ul><ul><ul><li>Easiest evolution of existing expertise and practices </li></ul></ul><ul><li>Why Platform? </li></ul><ul><ul><li>Maximizes component re-use </li></ul></ul><ul><ul><li>Minimizes hand coding </li></ul></ul><ul><ul><li>Maximizes flexibility and control </li></ul></ul>PaaS IaaS Built by user Provided by IT Built by user Provided by IT IaaS PaaS
What: Oracle Cloud Platform for PaaS Platform as a Service Infrastructure as a Service Oracle VM for x86 Operating Systems: Oracle Enterprise Linux Cloud Management Oracle Enterprise Manager Configuration Mgmt Lifecycle Management Application Performance Management Application Quality Management Database Grid: Oracle Database, RAC, ASM, Partitioning, IMDB Cache, Active Data Guard, Database Security Application Grid: WebLogic Server, Coherence, Tuxedo, JRockit Shared Services Integration: SOA Suite Security: Identity Mgmt Process Mgmt: BPM Suite User Interaction: WebCenter Oracle Enterprise Linux Oracle Solaris Oracle VM for SPARC (LDom) Solaris Containers Servers Storage Physical and Virtual Systems Management Ops Center Oracle Applications Third Party Applications ISV Applications
Private PaaS Lifecycle 1. Cloud Set Up 2. App Set Up 3. App Use 4. App Admin Self-Service Interface Shared Components Set up PaaS Set up self-service portal Set up shared components Dept App Build app using shared components Central IT Department App Owner Deploy using self service App Users App Owner Use app Oracle VM Oracle Enterprise Linux Oracle Database Oracle Fusion Middleware Oracle Enterprise Manager Manage app Adjust capacity Review chargeback
Evolving From Silos to Grid From Physical to Virtual <ul><li>Physical, dedicated silos </li></ul><ul><li>Sized for peak load </li></ul><ul><li>Difficult to scale </li></ul><ul><li>Expensive to manage </li></ul><ul><li>Virtualized, shared resources </li></ul><ul><li>Improved utilization </li></ul><ul><li>Scale as needed </li></ul><ul><li>Efficient to manage </li></ul>
Grid Computing: Virtualization & Clustering Cloud Is Not Just Server Virtualization
Sharing and Consolidation with Grid Computing Server A Server B Server C Server D Application A Application B Application C Application D Workload Avg Utilization <20% Applications A, B, C, D, E Net Workload Avg Utilization 70% Freed capacity to deploy elsewhere <ul><li>Take advantage of complementary workload peaks </li></ul><ul><li>Higher utilization rates and efficiency </li></ul><ul><li>Lower CapEx & OpEx </li></ul><ul><li>Green footprint </li></ul>Oracle Shared Instance Server E Application E Server A Server B Server C Server D Server E Virtualization and clustering enable consolidation
Elastic Scalability with Grid Computing Applications A, B, C, D, E Net Workload If utilization too high, increase capacity <ul><li>Pay-as-you-go scale-out </li></ul><ul><ul><li>Lower upfront CapEx and ongoing OpEx </li></ul></ul><ul><ul><li>Green footprint </li></ul></ul><ul><li>Rightsized capacity planning </li></ul><ul><ul><li>Smaller, standard machines running at higher utilization </li></ul></ul><ul><li>Defer equipment procurement </li></ul><ul><ul><li>Exploit advances in hardware price-performance and energy efficiency </li></ul></ul>Oracle Shared Instance Server A Server B Server C Server D <ul><li>World-class clustering at all levels: database, middleware, storage </li></ul><ul><li>Add/Remove nodes on-demand </li></ul><ul><li>Scale out as workload increases </li></ul>Scale-out on-demand
Quality of Service with Grid Computing Applications A, B, C, D, E Net Workload <ul><li>Systematic high Quality of Service </li></ul><ul><li>Reliability through redundancy </li></ul><ul><li>Predictable performance at any scale </li></ul><ul><li>High availability – every application gets HA </li></ul>Oracle Shared Instance Server A Server B Server C Server D <ul><li>Load balancing </li></ul><ul><li>Failover </li></ul><ul><li>Active-Active operation </li></ul>High performance and availability Server E <ul><li>Disaster recovery </li></ul><ul><li>Rolling upgrades </li></ul>
Grid Computing in All Tiers <ul><li>Middleware </li></ul><ul><li>Application Grid </li></ul><ul><ul><li>WebLogic Server </li></ul></ul><ul><ul><li>Coherence In-Memory Data Grid </li></ul></ul><ul><ul><li>JRockit Real Time </li></ul></ul><ul><ul><li>Tuxedo </li></ul></ul><ul><li>Database </li></ul><ul><li>Real Application Clusters </li></ul><ul><li>In-Memory Database Cache </li></ul><ul><li>Sun Oracle Database Machine </li></ul><ul><li>Storage </li></ul><ul><li>Automatic Storage Management </li></ul><ul><li>Exadata Storage Server </li></ul><ul><li>Infrastructure </li></ul><ul><li>Oracle VM </li></ul><ul><li>Oracle Enterprise Linux </li></ul><ul><li>Management </li></ul><ul><li>Oracle Enterprise Manager </li></ul>
Oracle IT: Oracle Development Self-Service Private Cloud Self-Service Application Job Mgmt Virtualization Priority Match Making Resource Mgmt Enterprise Manager Grid Control Submit Notifications Developer Metadata / Label Servers Results Hosts
Oracle IT: Oracle Development Self-Service Private Cloud <ul><li>Implementation Overview: </li></ul><ul><ul><li>Scope/Scale - Over 2600 physical servers with over 6000 Virtual Servers used by over 3500 developers </li></ul></ul><ul><ul><li>Activations – Processing over 70 jobs per day, this translates into over 45,000 jobs processed supporting production and test requirements. </li></ul></ul><ul><ul><li>Utilization – Rates on these servers averages 80% 7 days a week and can reach 90% during peak times. </li></ul></ul><ul><li>Results/Benefits : </li></ul><ul><ul><li>Increase in development productivity </li></ul></ul><ul><ul><li>Self-Service system for creation of development environments </li></ul></ul><ul><ul><li>Cleaner code lines as environments are created quickly for more thorough testing/validation. </li></ul></ul><ul><ul><li>Physical Server/Environmental Reduction by 75% </li></ul></ul><ul><ul><li>Server/Apps Deployment reduced by 80% </li></ul></ul>
Oracle IT: Oracle University Dynamic Provisioning with Grid Computing <ul><ul><li>Education Services </li></ul></ul><ul><ul><li>2,300 environments automatically provisioned weekly </li></ul></ul><ul><ul><li>1/10 th the hardware </li></ul></ul><ul><ul><li>CPU utilization increased from 7% to 73% </li></ul></ul><ul><ul><li>Floor space reduced 50% </li></ul></ul><ul><ul><li>Power consumption reduced 40% </li></ul></ul><ul><ul><li>Servers: Administrator ratio increased 10X </li></ul></ul><ul><ul><li>Revenue/Server increased 10X </li></ul></ul>
Oracle Cloud Computing Strategy Public Clouds IaaS PaaS SaaS I N T R A N E T Private Cloud Users IaaS PaaS SaaS I N T E R N E T IaaS PaaS IaaS PaaS SaaS SaaS Oracle Private PaaS Oracle Technology in public clouds Oracle Applications On Demand Oracle Applications
Oracle in Public Clouds <ul><li>Oracle Database, Fusion Middleware & Enterprise Manager supported on EC2 </li></ul><ul><li>Amazon Machine Images (AMIs) </li></ul><ul><li>Oracle Database Secure Backup to S3 </li></ul><ul><li>Self-service Public PaaS based on Oracle VM, Oracle Enterprise Linux, Oracle Database RAC and Oracle WebLogic Server </li></ul>
Oracle Applications Deployed on Shared Services Private PaaS Shared Components Industry Applications Private PaaS Oracle VM Oracle Enterprise Linux Oracle Database Oracle Fusion Middleware Oracle Enterprise Manager
Oracle On Demand Flexible Deployment Options Remote Management Hosted & Managed Multi-Tenant SaaS Single-Tenant SaaS On-Premise Pay-per-use Licensed OpEx CapEx & OpEx Off-premise On-premise Managed by vendor Managed by Customer Vendor scheduled maintenance Customer scheduled maintenance Public Private
Oracle Cloud Computing Strategy <ul><li>Oracle provides most complete, open and integrated cloud vision, strategy and offerings in the industry </li></ul><ul><li>Cloud is the evolution of capabilities Oracle has been working on for more than a decade: grid computing, virtualization, shared services and management systems </li></ul><ul><li>Oracle offers: </li></ul><ul><ul><li>Technology to build private clouds or run in public clouds </li></ul></ul><ul><ul><li>Applications deployed in private shared services environment or via public SaaS </li></ul></ul>