Published on November 23, 2009
Research Cloud Computing without the hype; an executive guide Defining cloud computing and identifying the current players Author: Steve Craggs Version 1.00 Page 2 November 2009
Table of Contents Executive Summary .................................................. 2 What is a cloud? ....................................................... 3 Why use a cloud? ..................................................... 4 Cloud flavours........................................................... 4 Cloud location and access options ..................... 4 What does the cloud offer, and in what form? ..... 6 The Cloud Computing marketplace .......................... 8 Some sample scenarios ........................................... 9 Gmail................................................................... 9 Amazon EC2 ....................................................... 9 Windows Azure ................................................... 9 Salesforce.com and Force.com ........................ 10 Shared testing facilities ...................................... 10 Storage clouds .................................................. 10 Summary ................................................................ 11 Glossary ................................................................. 12 Appendix-The Cloud Computing Market ................ 13 Infrastructure Services ....................................... 14 Platform Services .............................................. 14 Software Services ............................................. 15 Cloud Software ................................................. 16 Page 3
Disclaimer Whilst reasonable care and skill has been taken by Lustratus Research Limited (the company) in the preparation of this report no liability is accepted by the company (except in the case of death or personal injury caused by the company's negligence) by reason of any representation or any implied warranty condition or other term or any statutory or common law duty or otherwise howsoever arising for any direct or indirect general special or consequential damages or loss costs expenses or other claims (whether caused by the negligence of the company or otherwise) which come out of the provision of this report or its use. All trademarks are acknowledged as the property of their respective owners. About Steve Craggs Steve Craggs has been in the IT industry for more than 25 years, much of that time in IBM. Steve held various positions in the IBM development organization, culminating in becoming the Worldwide Executive in charge of IBM’s MQSeries middleware, now renamed as WebSphereMQ. Steve has been an independent consultant and analyst for the last 7 years, and is recognized as one of the leading authorities in the world on business integration software and solutions, and middleware in general. Steve is President of Saint Consulting, a Director of Lustratus and Vice-Chairman of the Integration Consortium, a global not-for-profit group for furthering the understanding of integration. Steve is a regular speaker at shows, events and webcasts in the US and Europe, and is a frequent contributor to a wide range of global IT publications. Steve publishes research and reports at http://www.lustratus.com/ and blogs on infrastructure technology at http://www.lustratusrepama.com/litebytes. Page 1
Executive Summary Executives are constantly faced with new initiatives from the IT world, bombarded by vendors, media and even their own internal teams with the latest buzz-words and IT innovations. In the face of this onslaught it can be really hard to sort out the wheat from the chaff, particularly since the IT market has a nasty habit of seizing on every new idea and trying to paint it as the latest ‘must have’ concept. Cloud Computing is the most recent in a long series of confusing new initiatives to emerge. Suddenly, everyone is talking about ‘The Cloud’, and banner headlines are trumpeting the fantastic savings / improved agility / better IT business alignment being offered. Executives find themselves struggling to know what to do, under pressure from myriad vendor claims and with a growing feeling of unease that somehow their companies may be missing out. The problem is that the IT industry is never very good at defining its latest terms, and for people without the time to get thoroughly immersed in the subject, or the persistence to cut through the hype, it can be almost impossible to make any sort of informed decision. This paper is designed to clarify some of this confusion, not at an in-depth technical level but from 50,000 feet – or 10,000 at least. The idea is to help the beleaguered executive get a reasonable grounding in Cloud Computing and its concepts, in order to support more accurate decision making and the optimal corporate IT strategy. A number of the main acronyms and terms will be covered, together with the main Cloud Computing operational models, and the Appendix provides a quick summary of some of the many vendors now playing in this marketplace. In the end analysis, there does appear to be something of value for most companies in Cloud. However, the key is to get enough understanding of the various flavours of Cloud and the different benefits promised by each in order for individual organizations to identify their best fits. Page 2
What is a cloud? There are many conflicting definitions of Cloud Computing in the industry, with different spins being chosen based on the specific agendas of the sources involved. However, at its heart the concept of Cloud Computing is based on a few fundamental precepts. •The cloud is a shared pool of IT resources – eg applications, Shared resource processors, storage, databases •Users can call up resources from the cloud, use them and then On‐demand release them when they are done in a self‐service fashion •More and more resource can be dynamically procured to satisfy Elasticity peaks in demand, and then released when demand subsides •The cloud is widely accessible, most commonly through the Networked access Internet Usage‐based •Cloud services are metered and accounted for at a usage‐based level metering Figure 1: Fundamental characteristics of Cloud Computing Those people familiar with data centre virtualization will relate immediately to the first point, and also to the second to some extent. This is what virtualization is all about – for instance, consolidate a bank of poorly utilized servers onto one large computer, where each group of users can be allocated a virtual environment to replace its original server. It will therefore come as no surprise that every Cloud Computing implementation has an underlying virtualization layer. However, Cloud is more than just virtualization. Usage-based scalability is one of the main areas of difference between the two concepts. While the virtualization concept allows a limited amount of dynamic scalability for a particular workload, the extent of this is bounded by fairly static configuration definitions. It is almost as if a workload is allocated a partition of the resources, and exceeding that partition requires IT specialists to evaluate and implement the change. The cloud concept demands extreme scalability. If more resources are needed than are currently available, Cloud Computing offers a self-service approach where users can quickly request additional capabilities or resources and these requests can be satisfied dynamically without waiting for static configuration changes. The final fundamental tenet of a cloud is that access should be widely available. Although this may sound like the recipe for anarchy, clouds include user-based metrics to ensure that the cloud provider can account individual resource usage to the appropriate user. And, of course, security mechanisms are used to limit access to authorized users only. Notice that the Cloud Computing definition does not specify exactly what resources are offered by the cloud, or whether the cloud is located within the corporate boundary or off-premise. These points reflect different flavours of Cloud Computing, and only serve to confuse the fundamental definitions – indeed a lot of the conflicting definitions of Cloud Computing result from falling into the trap of trying to define the overall concept based on one of the various different implementation models. These different versions of Cloud will be discussed later. Page 3
Why use a cloud? Before looking at the many different forms of Cloud Computing, it is worth considering the main advantage areas. The extent of these benefits will depend to some degree on the particular cloud implementation, but the basic principles of Cloud Computing offer some generic benefit areas that apply to all clouds. There are two primary areas of Cloud Computing benefits: Business delivery benefits o IT resources can be leveraged optimally to serve fluctuating business needs Users don’t hoard resources when they are not needed Users can get more resource dynamically, as required o Improved agility due to less reliance on the IT department to support and deliver new initiatives o New resource deployments can be delivered much faster o Resources can be accessed wherever and whenever required Infrastructure management benefits o Cost savings in designing, provisioning and delivering new services o Cost elimination through standardization and reuse of IT infrastructure o Less opportunity for costly human errors, for example in capacity planning for a new service o Usage based vs capital expenditure financial model (hardware and software) o Commitment to a ‘green’ corporate environmental strategy through efficient resource use o Access to cheaper sources of processing power The market hype surrounding Cloud Computing is intense, and executives should be suspicious about overly dramatic claims. For example, saying that Cloud Computing will eliminate the need for any internal IT investment is grossly overstating the case. However, the two prime benefit areas described offer a pretty compelling reason to take a closer look. The business delivery benefits stem from the self-service nature of cloud usage combined with the improved resource availability, efficiency and scalability. The infrastructure benefits come from the added efficiency of using a shared resource pool, the opportunity to outsource some IT management tasks and the metered nature of cloud usage. For example, a usage-based charging mechanism might be used to reduce software and hardware capital expenditure when using an externally provided cloud, or in the internal cloud case to implement a more acceptable accounting model where departments ‘pay’ for their IT services based on how much they use them. However, these benefits will be greatly affected by the choice of Cloud Computing model. It is the different flavours of Cloud that cause much of the confusion for executives trying to get their heads around the concept. Cloud flavours Cloud versions tend to differ across two different axes. Where are the cloud resources located, and who else has access to them? What IT resources / services are offered by the cloud, and in what form? By dealing with each in isolation, Cloud Computing should become a lot clearer. Cloud location and access options Contrary to the impression given by some, Cloud Computing is not synonymous with offloading IT workloads outside the enterprise into the hands of a Goggle or an Amazon. In fact, there are a number of options covering the way a cloud is deployed and accessed. The key variables are whether to have the cloud resources on- Page 4
premise or off-premise, and whether to restrict access to one organization or allow others to use the same cloud. The main trade-off here is the opportunity for cost savings versus the added potential for risk. Obviously keeping resources on-premise with internal use only access minimizes any security concerns, while utilizing an off-premise cloud that is open to all introduces the opportunity for economies of scale and a reduction in local IT staffing and licensing costs. Basically, the list below summaries the main Cloud Computing categories for location and access:- The public cloud – IT resources and services are owned by a third party, located off-premise and made available to anyone on a commercial basis as metered services The community cloud – IT resources and services are owned and operated on behalf of a community of organizations The private cloud – IT resources and services are owned/leased by a single company for its own use o The internal cloud – A private cloud where all resources remain on-premise The hybrid cloud – A combination of two or more cloud models Before looking at each category, one enormous source of confusion in the mayhem surrounding the emergence of Cloud Computing is the distinction between private and internal clouds. Often these terms are used interchangeably, since by definition an internal cloud is dedicated to the company concerned and is therefore ‘private’. However, purists are quick to point out that technically it would be possible for an organization to arrange for a private cloud to be implemented off-premise by a third party, or even to lease an off-premise cloud. In contrast, the term ‘internal cloud’ is used to imply not just a cloud dedicated to a single organization but one that is also implemented within the corporate boundary. Looking at each of these cloud options in turn, consider first the public cloud model. Here, because the resources within the cloud are not in the corporate domain, the consuming company does not have to worry about capacity management and associated investments in new hardware and other IT resources, or systems support activities. These responsibilities move to the public cloud provider. It may also be possible for the using company to avoid paying some software license fees – indeed this is a major incentive to many companies for using a public cloud. The metered, usage-based characteristic of Cloud Computing raises the possibility of a consumer paying a per-user / per-service usage charge instead of having to purchase licenses for individual software products. However, running all or part of commercial workloads in a public cloud will entail having to move some corporate data into the cloud, and this is one of the prime sources of risk in the public cloud model. The other risk most commonly encountered is based on the fact that in a public cloud model there will be many different using companies sharing the same IT resources. Although cloud providers make every effort to maintain security and integrity between these different ‘tenants’, some organizations consider the possibility of accidental interference or even security breaches to be a major problem, especially given that other tenants might be competitors. Additional risks include the possibility that the cloud provider will not adequately manage capacity, or even that some of them may fail commercially, leaving the user in trouble. Before moving on to other cloud models, there is another point that needs clearing up with regard to the tenants concept. Some public cloud suppliers will claim to offer private cloud facilities too. For example, they will point to their own security software that maintains the walls between the operations of different cloud tenants, and suggest that this produces the same result as a private cloud. The key to understanding the weak point in this argument is that in the private cloud model, the cloud is owned (or at least leased) by the consuming organization, and therefore the IT resources are totally dedicated to use by this single organization. In the public cloud case this is not the case. The consuming organization’s data and applications will be running within the domain of the cloud provider, which some organizations would regard as a security concern, and resources are allocated from and returned to a common pool which also represents a potential security issue. The community cloud offers an attractive option for a group of business partners wanting to share IT costs and improve efficiency across the community. The cloud location may still be external, perhaps controlled by a jointly-owned subsidiary, or may be on-premise at one or more partners. However, access is restricted to Page 5
community members, and therefore there is less concern over the exposure of data and the commercial viability of the supplier. Obviously, the security issues do not go away entirely, since even between partners there will be sensitive data that must be kept secret, but the risk is mitigated considerably. The private cloud almost entirely mitigates the cloud-related security concerns. In the internal cloud case where the private cloud is on-premise, security is very much business as usual. Concerns over cloud provider longevity vanish, but of course on the other hand the opportunities for cost reduction are much more limited. If the private cloud is off-premise this increases risk slightly, but no more so than moving to any form of outsourced arrangement. Some cost reduction is still possible however; efficient use of IT resources across the enterprise coupled with the user-oriented self-service nature of Cloud Computing combine to provide opportunities for reductions in IT support staffing for example, as well as improving service delivery. On the down side, on- demand scalability will be bounded by the enterprise’s own IT infrastructure, as opposed to having the opportunity to leverage the potentially larger pool offered by a public cloud provider. Finally, the hybrid cloud represents the idea of using two or more different cloud models. As Cloud Computing usage expands, it is generally accepted that most companies will end up using the hybrid cloud approach. One common reason given is that while public cloud might be the answer for some parts of the business that are relatively low risk, many companies will be reluctant to move sensitive data and applications outside the corporate boundary. In these scenarios, if a company has elected to move to a Cloud Computing model, it will be using both an internal and public cloud. However, the hybrid cloud definition is yet another area where confusion abounds. The best way to clear up this confusion is to start from the end-user perspective. In a hybrid cloud model, the end-user should perceive a single cloud, without any need to know where the resources being used reside. However, the confusion stems from the fact that there are different implementations of a hybrid cloud under this single-view interface. In some scenarios, the internal IT organization will want to force cloud selection based on the particular operation being executed – using the internal cloud for sensitive applications, for instance, and the public cloud for lower risk ones. In others, a level of portability will be desired to enable cloud selection to be dynamic, based on factors such as load-balancing and resource availability. An example might be an organization using a private cloud by default but taking advantage of public cloud resources to handle demand spikes. However, this approach requires the execution component and accompanying data to be portable between the internal cloud environment and the public one. What does the cloud offer, and in what form? Potential cloud users will hear all sorts of terms thrown around – storage clouds, compute clouds, SaaS, desktop clouds and so on. However, all these variations split into three prime cloud delivery models; Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS). Even this terminology is confusing, since it implies that these delivery modes are all offered to the cloud user as services, and in the IT industry ‘services’ is a heavily over-burdened term. It is best not to get too hung up with the wording, but simply to think of the acronyms, IaaS, PaaS and SaaS. In essence, these three terms balance flexibility against complexity. Page 6
IaaS, PaaS and Saas ‐ Complexity versus flexibility of use Infrastructure as a Service Complexity Platform as a Service Software as a Service Use case flexibility Figure 3: Cloud delivery models – how use case flexibility drives complexity IaaS describes a model where the cloud delivers an IT infrastructure – some combination of hardware and system software - to each user. It is left to the user to then pull together the requisite infrastructure and development tooling to build the applications that will make use of this infrastructure. Think of a department asking for some new computer equipment. Instead of submitting a request for the procurement department to go and buy the desired hardware and deliver it, the user department request the resources from the cloud. PaaS is less flexible than IaaS, but is also less complex. In the PaaS model, a user requests a platform on which its applications can be built and then run. This platform will already be set up with development products and tools, or perhaps may even support standard development environments such as Java or .NET. The user still has to build the applications that will run on the cloud resources, but a lot of the lower level implementation decisions have been taken. Think of a department being supplied with the standard server configuration supported by the enterprise, for example a LINUX server with IBM’s WebSphere Application Server and DB2 database system, and a WebSphere-based development toolset. The SaaS model is the least complex but also the most narrowly focused. Here, the user requests to use a particular application in the cloud. This is typically pre-packaged and offers little opportunity for customization. Google Mail is an example of this type of SaaS model. The table below summarizes the characteristics of these three main Cloud Computing deployment options. Page 7
Delivery model Use case flexibility Options Infrastructure The cloud delivers The user can choose how Options might include choice as a service combinations of hardware these resources are used of operating system, (IaaS) and system software application server, database and packages Platform as a The cloud delivers a platform The user writes his own The user may also bring in service for building and/or running applications to whatever third party applications (PaaS) applications interface the cloud platform written for the cloud platform supports Software as a The cloud delivers software The user accesses and runs More applications can be service applications as services the applications as provided selected for additional fees (SaaS) Figure 4: Cloud Computing deployment options The Cloud Computing marketplace Having looked at what Cloud is, and how it is deployed and/or consumed, it is worth spending a little time on the Cloud Computing marketplace. This is another source of confusion for many cloud novices. Going back to basics, Cloud Computing requires a virtualized pool of IT infrastructure, a layer of software to provide the Cloud Computing functionality and cloud-enabled use cases to deliver the business value. SaaS PaaS IaaS Cloud Use Case Cloud supplied User/ISV developed Infrastructure building applications applications blocks Cloud Software Cloud Software Secures, scales, optimises, manages and charges for virtual infrastructure Virtualisation Virtualisation Software Software Presents virtual infrastructure to the layers above Physical Physical Infrastructure Infrastructure Underlying Hardware, System Software and Network Figure 5: Cloud Computing architecture The cloud implementation layer of software in the diagram above forms the cloud software market segment, and it is this software that a company would buy in order to build its own internal cloud. Suppliers of public clouds, or ‘cloud providers’, use the same type of cloud software internally, but then offer access to their clouds in the form of cloud services. The question over who builds the cloud-enabled applications is resolved by the choice of cloud delivery model – in the SaaS case, the applications are provided by the cloud provider or other ISVs while in the PaaS and IaaS cases the applications are built by the using company. Therefore, the ‘services’ provided by cloud providers may range from delivering infrastructure such as archive storage to providing ready-made business applications to access and run. Page 8
To summarize, cloud software vendors sell software used to build and operate cloud environments, while cloud providers sell services sourced from cloud environments that they have built themselves. An important aspect of these two different approaches in commercial terms is the way the user justifies and pays for the cloud. In the internal case, this is just like a traditional software implementation, where software licenses are acquired and an ongoing maintenance agreement is put in place. But in the external cloud case, the cloud provider bears all of these costs, and balances this by charging a usage-based fee – often in per user / per service / per month terms. This is one of the main reasons why the public cloud approach offers the best opportunities for cost reduction. Of course in the internal cloud case, because of the usage-based nature of Cloud Computing the end-user departments may still be charged for IT usage in the same fashion as for a public cloud, but this is purely an internal accounting mechanism. Another point to note about the marketplace is that Cloud Computing assumes the presence of a virtualization layer to make the hardware pooling aspect work. Some Cloud Software vendors also offer the virtualization software, while others rely on the user already having a virtualization layer in place. Some sample scenarios Having discussed what Cloud Computing is, in its many different forms, it may be helpful to consider a few sample scenarios to show how the definitions work and how Cloud might be used. Gmail Google’s email offering, Gmail, has many users across the world. Users set up a Gmail email address, and then use the Gmail email software offered by Google through the web browser. Gmail is free, funded by the fact that users are subjected to advertising when they use the software. This is a classic Cloud Computing scenario. Much of the user interface is self-service – users set up a mail ID, and can request additional mailbox or archive storage as required, although beyond a certain level this extra storage is not free. In the deployment model terms, this is SaaS – the applications are offered pre-packaged. It is also a public cloud implementation, since the applications and storage live in the Gmail cloud and not on the user’s site. Amazon EC2 The Amazon Elastic Cloud Compute facility allows users to access computer power on demand. It does this by allowing consumers to create their own workload ‘image’; consisting of the data, applications, libraries and configuration specifications required to execute the desired business functionality; and then deploy this to run on hardware in the Amazon Cloud. Amazon already invests heavily in data centres to power its own businesses and it is their expertise in managing this computing resource that they are “re-selling” in this way. There are all sorts of options available, such as what operating system to use, how much storage and memory will be needed and how many instances of the application to run. Amazon claims that all you need to have your applications execute in their cloud is a credit card and an Internet connection. This is an example of a public cloud delivering IaaS – the Amazon Cloud provides the hardware and the system software as requested, and the user delivers everything that is to run on that infrastructure. Windows Azure Microsoft has its own Cloud offering, the Windows Azure Platform. Once again, this is a public cloud implementation, with the IT resources being provided by Microsoft data centres around the world. Azure is an operating system that overlays the IT infrastructure. The Azure cloud user is provided with a library of tools for building the desired Azure applications using popular tools such as Visual Studio. This is an example of PaaS, where the user is provided with the environment but must create the applications to be deployed in the cloud. Page 9
Salesforce.com and Force.com Saleforce.com has been selling hosted sales force automation applications for a number of years. Users are charged on a per user / per month basis, depending on which Salesforce.com modules they need. This is a SaaS model deployed in a public cloud. However, it is worth mentioning because Salesforce.com also offers Force.com, which is its PaaS offering. The idea is for other organisations, both enterprises and independent software vendors, to build applications to run in the Salesforce.com cloud, and therefore a platform is required to achieve this. Force.com is also used by Salesforce.com users who want to customize the base applications. Shared testing facilities One IaaS scenario that is proving quite popular is the idea of getting multiple different test departments to share IT resources from a pool, instead of each having to acquire its own infrastructure. This offers efficiency gains and also saves a lot of configuration and support time. In some cases this scenario has been deployed as an internal, or private cloud, where the users are different departments within the enterprise. However it has also been used very successfully in a community cloud model. For instance, one example was a government- backed scheme to stimulate small company activity, providing computing resources for these companies without the need to incur capital expenses until the commercial model had been developed. Storage clouds This concept is getting quite a lot of attention too. Many companies are struggling with the sheer volume of data they have, much of which is rarely used and is a good candidate for archiving. Every time a user asks for more storage space, the danger is that every allocation will have to have a buffer of free space, and hence the efficiency of storage usage can be poor. A cloud, whether internal or external, provides a convenient way to manage this storage efficiently. This, of course, is an example of IaaS. Page 10
Summary Cloud Computing is an exciting new area, and does seem to merit investigation for many companies. While there is a tremendous amount of confusion over Cloud, once the high level view has been grasped it becomes a lot easier to put the various market claims into perspective. The diagram below summarizes the key points about the Cloud Computing marketplace, as discussed in the previous sections. Deployment Models Service Models Private / Internal cloud Community cloud Cloud Software as a Service (SaaS) Essential Characteristics Public cloud Cloud Platform as a Service (PaaS) Broad On‐demand Resource Rapid Measured network self‐service pooling elasticity Service access Cloud Infrastructure as a Service (IaaS) Hybrid cloud Figure 6: The Cloud Computing market landscape The key issue that executives must address with relation to Cloud Computing is to understand the company attitude to risk, and to balance this against the benefits expected from adopting a cloud model. For instance, many companies considering cloud are worried about the potential security risk with public clouds, no matter how much the cloud providers trumpet their security and integrity measures. While these risks may be acceptable for low-impact areas like user productivity tools, commercial business applications are another matter. There are also very real concerns about maturity levels in cloud development and technology, and for some IT organizations even the concept of IT as a service feels uncomfortable. However there is considerable interest in exploring the possibilities of cloud-based infrastructure, and as a result, a lot of companies are looking to deploy Cloud Computing in a limited fashion, internal-only and for a restricted subset of operational activities. This makes a lot of sense – try it out in a low-risk scenario and validate the benefit claims before moving to more adventurous deployments. It is this approach of plotting a slow and steady course towards Cloud Computing rather than rushing headlong into the brave new world that is being adopted by most businesses today. Page 11
Glossary Private Cloud A cloud that is owned or leased by a single consuming enterprise Public Cloud A cloud that anyone can use Community Cloud A cloud that is operated on behalf of a community of companies Internal Cloud A private cloud that is located on-premise Hybrid Cloud A cloud that utilizes some combination of public, private and community clouds Cloud A pool of IT resources such as processors, storage, systems software, networking and applications that can be shared between different users, with the following characteristics: • Users can call up resources from the cloud, use them and then release them when they are done in a self-service fashion • More and more resource can be dynamically procured to satisfy peaks in demand, and then released when demand subsides • The cloud is widely accessible, most commonly through the Internet • Cloud services are metered and accounted for at a usage-based level Cloud provider A commercial organization owning a cloud and selling usage-based services resourced from it Cloud Software The software required to implement a cloud. The cloud software layer usually assumes the presence of a virtualization layer Cloud Software vendor A commercial organization selling software used in building and operating a cloud Usage-based pricing The typical form of commercial pricing for public cloud services, eg per user / per hour / per service Virtualization software The software required to turn a collection of IT resources into a shared pool, where resources are allocated depending on user needs Page 12
Appendix-The Cloud Computing Market This appendix lists some of the many players in the Cloud Computing marketplace. Lustratus is tracking a large number of commercial organizations providing cloud services, selling cloud software or offering cloud-related professional services, and a detailed examination of each is far beyond the scope of this introductory paper. Instead, this section attempts only to provide a summary of which vendors play in what part of the overall Cloud Computing market. At the least this should give potential adopters of Cloud Computing a starting point when looking to find the most appropriate suppliers. Recapping on the earlier analysis in this guide, the basic categories of suppliers are Cloud suppliers, offering services sourced from their own clouds o Software Services (Software as a Service, providing access to applications) o Platform Services (Platform as a Service, providing a place to run developed applications) o Infrastructure Services (Infrastructure as a Service, providing access to hardware and system software) Cloud-related professional service suppliers, to aid in implementing and using Cloud Computing Cloud software vendors, selling software for implementing and using your own cloud The diagram below illustrates the market segmentation across these broad categories used by Lustratus in its Cloud Computing taxonomy. This segmentation is discussed in more detail in the Lustratus Research paper, “Cloud Computing – Market Landscape”. Software Platform Infrastructure Professional Cloud Services Services Services Services Software Billing, Financials, Legal, Strategy, planning and Human Resources* General Purpose design Data Storage Sales, CRM Migration and implementation services Desktop Productivity Business Intelligence Compute Testing consultancy Content Management Integration Compute Cloud Management Backup & Recovery Security consultancy Document Management Development & Testing Appliances Cloud application design/porting Collaboration Services Management Database Support services and File Storage Social Networks training Figure 7: Technical segmentation of the Cloud Computing market The least well developed area at the moment is that of professional services, with almost all current offerings being provided by vendors in other Cloud Computing categories. For the other categories, however, the following sections offer a list of vendors that Lustratus has identified as potentially relevant. Note that Lustratus is not giving any judgement on how well each vendor serves its segment – the list is purely a segmentation based on vendor claims and other publicly available material. Page 13
Infrastructure Services Sub-segment Working description Vendors Storage Storage services typically provide metered, Amazon EBS, Amazon S3, CTERA Portal, on-demand storage of structure or un- Mosso Cloud Files, Nirvanix, Vaultscape structured data delivered as a service Compute Compute services typically provide metered, Amazon EC2, ServePath GoGrid, Elastra, on-demand execution of code fragments, Mosso Cloud Servers, Joyent Accelerators, services or applications - delivered as a AppNexus, Flexiscale, Elastichosts, service Hosting.com CloudNine, Terremark, GridLayer, iTRICITY, LayeredTech, Zimory Public Cloud Services Services management services typically RightScale, enStratus, Scalr, CohesiveFT, Management provide a layer of management that sits Kaavo, CloudStatus, Ylastic, Dynect, above the underlying infrastructure services CloudFoundry, NewRelic, Cloud42 providing aggregation, scaling, execution management, fault tolerance and security of the underlying infrastructure services Platform Services Sub-segment Working description Vendors General General purpose platform services provide Force.com, Etelos, LongJump, AppJet, Purpose metered, and/or on demand development Rollbase, Bungee Labs, Connect, Google tools, frameworks and/or execution App Engine, Engine Yard, Caspio, Qrimp, environments that allow cloud-capable MS Azure Services Platform, Mosso Cloud applications or application fragments to be Sites, WorkXpress built, deployed and managed Business Business intelligence services provide Aster DB, Quantivo, Cloud9 Analytics, Blink Intelligence metered, and/or on demand business Logic, Analytics, LogiXML, Oco, Panaroma, intelligence and analytics capabilities that are PivotLink, Sterna, ColdLight Neuron, delivered as a service and can be cloud- InfoBright, Vertica capable or cloud-aware in that their data sources may be cloud-based. Integration Integration services provide metered, and/or Amazon SQS, MuleSource Mule on demand connectivity, integration and OnDemand, Boomi, SnapLogic, OpSource process orchestration capabilities delivered Connect, Cast Iron, Microsoft BizTalk (.Net) as a service and can be cloud-capable or Services, gnip, SnapLogic SaaS Solution cloud-aware in that the systems they Packs, Appian Anywhere, HubSpan, integrate can be cloud-based. Informatica On-Demand, TIBCO Silver, Cordys Development Development and testing services provide Keynote Systems, Mercury, SOASTA, and Testing metered, and/or on demand software SkyTap, Aptana, LoadStorm, Collabnet, development life-cycle management Dynamsoft capabilities delivered as a service. They can be cloud-capable or cloud-aware in that the development projects may themselves be cloud-based. Database Database services provide metered, and/or Google BigTable, Amazon SimpleDB, on demand structured data management FathomDB, Microsoft SQL Azure services to application developers delivered as a service. Page 14
Software Services Sub-segment Working description Vendors Billing Software services provide metered and/or Aria Systems, eVapt, OpSource, Redi2, on-demand access to specialised business- Zuora level applications, typically accessed by end- users and are delivered as a service. Financials Concur, Xero, Workday, Beam4d Legal DirectLaw, Advologix, Fios, Sertifi Human Taleo, Workday, iCIMS Resources Sales Xactly, LucidEra, StreetSmarts, Success Metrics CRM NetSuite, Parature, Responsys, Rightnow, Salesforce.com, LiveOps, MSDynamics, Oracle On Demand Desktop Zoho, IBM Lotus Live, Google Apps, Productivity Desktoptwo, Parallels, ClusterSeven Content Clickability, SpringCM, CrownPoint Management Backup & JungleDisk, Mozy, Zmanda Cloud Backup, Recovery OpenRSM, Syncplicity Document NetDocuments, Questys, DocLanding, Management Aconex, Xythos, Knowledge TreeLive, SpringCM Collaboration Box.net, DropBox Social Ning, Zembly, Amitive Networks Page 15
Cloud Software Sub-segment Working description Vendors Data Data management cloud software typically 10Gen MongoDB, Oracle Coherence, provides database-like functionality that can Gemstore Gemfire, Apache CouchDB, be deployed within a cloud-enabled Apache Hbase, Hypertable, TerraCotta, architecture possessing some of the Tokyo Cabinet, Cassandra, memcached following attributes: massively distributed, parallel data processing, massively scalable, clustered, cached and/or replicated. Compute Compute cloud software allows a virtual Globus Toolkit, Xeround, Beowulf, Sun Grid environment to be built into which Engine, Hadoop, OpenCloud, Gigaspaces, applications or application fragments can be DataSynapse, Platform, Zimory Enterprise deployed and executed. This is typically Cloud achieved by adding a management abstraction layer across a massively distributed network of hardware Cloud Cloud management software allows OpenNebula, Enomaly Enomalism, Altor Management enterprises and service providers to manage Networks, VMWare vSphere, CohesiveFT the components of their virtual cloud VPN Cubed, Eucalyptus, Reductive Labs infrastructure. These elements may include Puppet, OpenQRM network, compute, storage or applications*. Management capabilities may include provisioning, starting, stopping, monitoring, moving, scaling or securing a service within public, private or hybrid clouds Application A subset of the cloud management segment Appistry, Open.ControlTier, DataSynapse Services , application services management software Fabric Server (Now TIBCO), AppZero (was Management provides automated management of the Trigence), 3Tera AppLogic, SpringSource lifecycle of application services including (was Hyperic) provisioning, starting, stopping, moving, monitoring , scaling and billing. Appliances Appliances provide hardware-based Symplified, Vordel, rPath*, PingIdentity* acceleration support services for various cloud functions File Storage File storage cloud software provides EMC Atmos, ParaScale, Zmanda, CTERA enterprises that are building internal clouds or providers that are building external clouds with massively scalable file storage capability. Page 16
About Lustratus Research Lustratus Research Limited, founded in 2006, aims to deliver independent and unbiased analysis of global software technology trends for senior IT and business unit management, shedding light on the latest developments and best practices and interpreting them into business value and impact. Lustratus analysts include some of the top thought leaders worldwide in infrastructure software. Lustratus offers a unique structure of materials, consisting of three categories—Insights, Reports and Research. The Insight offers concise analysis and opinion, while the Report offers more comprehensive breadth and depth. Research documents provide the results of practical investigations and experiences. Lustratus prides itself on bringing the technical and business aspects of technology and best practices together, in order to clearly address the business impacts. Each Lustratus document is graded based on its technical or business orientation, as a guide to readers. http://www.lustratus.com/ Lustratus REPAMA Lustratus REPAMA is the marketing consultancy and competitive marketing intelligence division of the analyst firm Lustratus Research Limited. Whilst Lustratus Research provides detailed technology analysis, reports, insight and advice aimed specifically at end users of technology; Lustratus REPAMA helps technology vendors to better compete in their markets. The REPAMA research methodology is central to Lustratus’ consultancy services and provides a detailed map of the go-to-market strategies of the vendors in a particular market segment. We represent these strategies and tactics graphically as well as textually which makes it simpler to compare vendors’ strategies and to identify strengths and weaknesses. REPAMA is an acronym formed from the phrase Reverse-Engineered Positioning and Messaging Analysis. Lustratus REPAMA services clients worldwide from its base in the UK. http://www.lustratusrepama.com/ Terms and Conditions © 2009—Lustratus Research Ltd. Customers who have purchased this report individually or as part of a general access agreement, can freely copy and print this document for their internal use. Customers can also excerpt material from this document provided that they label the document as Proprietary and Confidential and add the following notice in the document: “Copyright © Lustratus Research. Used with the permission of the copyright holder”. Additional reproduction of this publication in any form without prior written permission is forbidden. For information on reproduction rights and allowed usage, email email@example.com. While the information is based on best available resources, Lustratus Research Ltd disclaims all warranties as to the accuracy, completeness or adequacy of such information. Lustratus Research Ltd shall have no liability for errors, omissions or inadequacies in the information contained herein or for interpretations thereof. Opinions reflect judgment at the time and are subject to change. All trademarks appearing in this report are trademarks of their respective owners. Page 17
Lustratus Research Limited St. David’s, 5 Elsfield Way, Oxford OX2 8EW, UK Ref STC/LR/33508620/V1.0 Tel: +44 (0)1865 559040 www.lustratus.com
Presentación que realice en el Evento Nacional de Gobierno Abierto, realizado los ...
In this presentation we will describe our experience developing with a highly dyna...
Presentation to the LITA Forum 7th November 2014 Albuquerque, NM
Un recorrido por los cambios que nos generará el wearabletech en el futuro
Um paralelo entre as novidades & mercado em Wearable Computing e Tecnologias Assis...
... on Cloud Landscape: Potentials and Paradoxes. ... www.slideshare.net/LustratusREPAMA/cloud-computing-without-the-hype-an-executive-guide-100 ...
d.pradnya3080 (d.pradnya3080) has uploaded 12 documents on Scribd.
Free Cloud Computing Readiness ... The last thing you want to do is build a cloud platform without direct ... • Executive Strategy • Cloud Computing ...
If you follow cloud computing, ... SECaaS, DaaS, MaaS and BaaS? Trending: IBM union calls it quits; ... Network World | Aug 13, 2012 1:00 AM PT RELATED ...
How the company got into cloud computing: ... want the benefits of cloud computing without the hassle of ... executive vice president of ...
IT strategy is a comprehensive plan that information technology management professionals use to guide their organizations.
Top 10 Cloud Myths. ... Cloud Computing; Future of IT Sales; Gartner Predicts; Information Security; ... Gartner CIO & IT Executive Summit.