CIW03082USEN

8/13/2019 CIW03082USEN

1/16

IBM Office of the CIO

Thought Leadership White Paper

July 2013

Success in the cloud: Why workload mattersObservations from IBMs own cloud transformation

8/13/2019 CIW03082USEN

2/16

2 Success in the cloud: Why workload matters

Contents

2 Executive summary

3 Introduction

3 IBMs impetus for adopting cloud

4 The importance of workload selection

5 Clouds biggest impact at IBM

12 IBMs other successful workload migrations

14 Clouds role in reinventing IBMs business

15 Conclusion

15 For more information

Executive summaryLike other performance-driven companies, IBM is continuouslychallenged by stakeholders to drive new revenue opportunities

and efficiencies while lowering operating costs. As such, cloudcomputing with its widely-touted benefits made a convincingcase for adoption. And the technology model has lived upto expectations. By providing a platform to standardize andautomate the management of key business applications, cloudhas enabled dramatic reductions in IBMs IT support costs andmajor improvements in workplace efficiency and resource use. More than that, cloud computing has accelerated IBMs ability toinnovate and do so in ways that improve how we deliver servicesand support inside IBM. In short, it has become a catalyst forbusiness transformation at IBM.

With the implementation of cloud computing internally, acrosssix fundamental IT workloadsdevelopment and test, analyticsstorage, collaboration, desktop and production application workloadswe have witnessed striking improvements inefficiency while capturing some impressive savings in capitaland operations. Consider that:

IBM development teams have seen server provisioning and

configuration drop from five days or longer to as little as onehour. IBMs development and test cloud has virtually elimi-nated IBMs testing backlog, speeding new development andenabling applications to reach the market sooner.

IBMs analytics cloud put an end to siloed business intelligenc(BI) and the six-figure funding required for new BI projects.Organizations across IBM are tapping into a centralizedanalytics cloud for tools and intelligence aggregated fromhundreds of information warehouses. The business value ofthese analytic insights is estimated to be more than USD $300million for the top 20 consumers of the service alone, out ofapproximately 300 active projects.

IBMs Virtual Storage Center cloud cut the cost-per-byte ofdata stored by nearly 50 percent at each of the IBM facilitiesin which it was implemented. This has allowed the facilityto accommodate the explosive growth in storage demand25 percent or more annuallywithout increasing its totalstorage budget, and it is expected to be able to do so for fourstraight years.

IBM Connections, our social software platform in the cloud,has dramatically increased workplace collaboration, productivity and innovation, while enabling IBM to locate and stayconnected with clients, and build awareness of our productsand services. The cloud currently supports over 50 million web conferencing minutes per month.

8/13/2019 CIW03082USEN

3/16

3IBM Office of the CIO

Through these and other internal cloud implementations,IBM is seeing firsthand the impact that cloud can have on thebusiness. There is little question about the potential of cloudcomputing to drive efficiency and lower costs. But the existing workloads a company selects for migration to the cloud andtheir affinity for the attributes inherent in the cloud model havea lot to do with clouds success as a transformative tool for thebusiness. As cloud matures, some of its most effective uses are

expected to come from workloads and new applications specifi-cally designed for cloud, so called born on the cloud workloads.

IntroductionIBMs IT leaders viewed cloud computing as an opportunityto radically simplify aspects of an IT operation that had growncomplex and less productive. Based on our own cloud researchand years experimenting with similar technology models, weunderstood the potential. Cloud could help IBMs developmentand testing teams provision their own server and storage capac-ity without week-long delays or involvement from systemadministrators. Cloud could help us move employees from aresource-straining traditional desktop environment to a virtualdesktop environment, facilitating new deployments, upgradesand end-user support. Cloud could facilitate online collaborationamong IBMs global workforce, making it easier for employeesto interact, share ideas and innovate with clients, businesspartners and each other.

Cloud has succeeded in doing all of these things, but muchhas been learned along the way. As with any new technology,cloud is best deployed in the right circumstances and with the

right workloads. One of the most significant determinants ofsuccess in moving to the cloud is the careful selection of those workloads. Some workloads are simply a better match for cloudcomputing, with more to gain from clouds intrinsic features.

This paper shares IBMs observations and recommendations with respect to workload affinity for cloud computing. Itdescribes the quantitative and qualitative value that IBM hasachieved with six common IT workloads. And while each ofthese workloads has benefitted from the cloud, they havebenefitted in different ways and to different degrees. Four work-loads, development and test, analytics, storage, and collaborationand social business workloads, have produced the most stunning

results to date. They have transformed the way whole groups ofIBM users do their jobs, enabling them to speed new develop-ment and uncover new sources of revenue, among other things. The results achieved with desktop and production application workloads, though not as pronounced, have been very promisingso far. As of this writing, they are still evolving and will continueto do so as their respective implementations mature.

IBMs impetus for adopting cloudIBMs intent in adopting cloud was not unlike any otherbusiness. With its sizeable prospects for controlling IT costs,enabling business and IT agility and delivering everything frominfrastructure resources to business processes as a service, cloudhad the potential to radically change the economics of IT.It could drive up the value that IT provides to the business whilelowering capital and operational costs. Cloud also offered themeans to deliver on the increasing demands of IBM employees,business partners and customers who have come to expect a newstandard of service.

More importantly, these cloud-enabled cost and operationalefficiencies had the potential to transform IBMs business, not just in how IT resources and services were delivered, but in how

IBM conducted business with clients and partners around theglobe. Cloud provided the innovative and collaborative platformand the computing flexibility to reinvent business at IBM.

8/13/2019 CIW03082USEN

4/16


The importance of workload selectionCloud computing has captured the attention of technology andbusiness leaders alike, but the actual value it delivers to theorganization varies with the application. When applied to theright workloads, cloud can deliver game-changing value. Whenapplied to the wrong workloads, clouds value over traditionaldelivery models can be diminished or lost altogether.

In selecting workloads for cloud, it is important to considerthe soft, intangible benefits as well as the more visible and easilyquantifiable benefits. After all, intangibles like customer satisfac-tion and quality of service are often significant elements of cloud value. For some workloads, they may represent the lions shareof the value delivered.

Clearly, some workloads have more to gain from a move to thecloud. This is often because they have a greater affinity for theattributes inherent in the cloud model. They align with the stan-dardization, virtualization, automation and level of managementand hardware support that a cloud service provides. These workloads can operate easily in a virtualized, automated cloudenvironment, where the infrastructure makes it possible todynamically request services from a virtualized pool of hardwareand then automatically provision the required software stack andresource capacity.

Existing workloads with the greatest affinity for cloud may bethose that are an excellent fit in terms of their potential gainand ease of deployment. These workloads require little to nocustomization because they can work with and benefit from theclouds standard catalog services. Figure 1 illustrates how several

common IT workloads fare on these two measures, based onIBMs research and experience with our own cloud implementa-tions. In general, the workloads that appear in the upper rightquadrant have proven to be the best fit for cloud computing.

Existing workloads must be carefully analyzed to weigh potentiagains against how easily they can be deployed in the cloud, whether public, private or hybrid. The fact is not every workloais the same in terms of its importance and cost to the organiza-tion, and this can affect its outcome in the cloud. Some of themost critical workloads are so costly to the organizationfinancially and operationally that a move to the cloud has thepotential to provide considerable benefit. Other workloads maybe so highly optimized already that there is little to be gainedfrom such a move.

H i g h

G a

i n

L o w

Low HighDeployment ease

Complex/criticalproduction

Simple/non-criticalproduction WebMail

(hosted)

Storage

Analytics

Collaboration

Development/ test

Self-serveapplication

development

environment

Complex ERP Complex/critical

production (virtualized)

Figure 1. Workload affinity for cloud computing. Potential for gain anddeployment ease are two key factors in the determination of a workloadsaffinity for and prospective success in the cloud.

8/13/2019 CIW03082USEN

5/16


Workloads that are unusually complex may offer big potentialfor improvement, but they may require a high degree ofcustomization or application redesign to conform to the cloudsarchitecture. These kinds of workloads can prove too difficult,risky or costly to move to the cloud. Certainly, the risk/rewardprofile of any workload should be clearly understood beforeit is deployed in the cloud. Consider complex mission-criticalproduction workloads. In some cases, the risk associated with

cloud deployment can far outweigh the potential gains. Similarly,legacy and heterogeneous applications, and workloads thatinvolve a high amount of data transfer, may be difficult to deployin the cloud, relative to their anticipated gain.

By contrast, self-contained applications may be easy to deploybut only offer moderate gains. Other workloads may be viewedas a good fit for cloud deployment simply because they poselittle risk to the organization from a security or other standpoint. This is the case with collaboration and social business workloads,for which there is little to lose and much to be gained from amove to the cloud.

In IBMs own cloud experience, we have seen each of theserisk/reward scenarios play out. But with cloud computing tech-nologies and practices continuing to advance, the risk/rewardprofiles associated with most cloud workloads are improving. This is especially true for workloads and applications thathave been specifically designed for cloud, often referred to asborn on the cloud workloads. Naturally, these applicationsand workloads are better able to exploit the services andcapabilities inherent in the cloud platform, so the payoff can beconsiderably larger. Moreover, as vendors add more capabilities

to their cloud platforms, born on the cloud applications willbe better able to leverage these functions to handle some of thetraditional heavy lifting of IT operations and delivery.

The remainder of this paper looks at the six internal IT work-loads that IBM migrated to clouds, starting with the four thathave had the greatest business impact to datedevelopmentand test, analytics, storage, and collaboration and social business workloadsfollowed by desktop and production application workloads. For all six, the paper explains why IBM elected tomove to the cloud and the benefits of doing so.

Clouds biggest impact at IBMIBMs greatest cloud computing gains to date have come frommigrating workloads that impacted the efficiency of key groupsof employees and ultimately IBMs ability to innovate. These workloads really stood to benefit from the change operationally,but also economically. On the one hand, they provided the mostopportunity for transformative improvement. On the other, they were tailor-made for the elastic resource allocation, provisioningspeed, extreme transaction processing and self-service automa-tion that are synonymous with cloud.

Development and test workloadsIBMs IT organization knew that cloud offered a better wayfor internal development teams to build and manage their testenvironments. After all, 30 to 50 percent of all IBM servers weretypically dedicated to test, and most were running at less than10 percent utilization. When developers requested access to theservers, they could expect to wait up to a week. So it was notsurprising that these teams would hold on to these resources when they were finally able to secure them, rather than releasingthem for the gaps of time they werent in use. The testingbacklog created by this and other provisioning challenges hadbecome the single biggest factor in the delay of IBMs new

application deployments.

8/13/2019 CIW03082USEN

6/16


Implementing the development and test cloud enabled theIT organization to overcome these issues. It provided infrastruc-ture as a service (IaaS with virtualized server, storage, operatingsystem and middleware) capability in support of the teams thatdevelop most of IBMs internal applications. It enabled IT toexpand these resources on demand, making new server imagesavailable as provisioning requests were made.

With the cloud, development and test teams saw their serverprovisioning and setup time drop from five days to as little asone hour. Instead of tapping systems administrators to manuallyprovision needed resources, they could provision resourcesautomatically and independently using a cloud service catalogcontaining standard operating system and middleware images. The catalog is regularly updated with new images to meet userdemand, and it simplifies chargeback with automated usagemetering and billing.

The images provide the standard build for the test environment.Since developers are no longer responsible for the builds, therisk of defects resulting from incorrectly configured test environ-ments (which stood at 30 percent prior to cloud) has lessenedsignificantly. The cloud not only lifted the configuration burdenfrom developers, it helped improve the quality of their testenvironments.

IBMs hosting support team also benefitted from the develop-ment and test cloud. They were able to realize substantial costsavings resulting from the clouds self-service and automationfeatures, which lowered users need for assistance. The cloudgave them considerably more time to focus on activities that

would lead to growth, new clients and innovations in technologyand services.

When IBMs development and test cloud was first deployed, we projected a 50/50 split between cloud-based self-provisioningand traditional manual provisioning requests. However, fromthe day the cloud-based option was introduced, it has been

embraced by the IBM development community, with thelarge majority choosing to go with the cloud. Today more than95 percent of server provisioning and de-provisioning requestsare made via the cloud, where it is available. This is a testamentto the speed, ease of use and streamlined process made possibleby the cloud.

Furthermore, de-provisioning requests are occurring at a much

higher rate with the cloud than they were with the manualprovisioning process. Clearly this is a function of the cloudsability to respond to provisioning requests in a matter of hours.Developers are now confident that capacity and resources can beacquired as needed, so they are more inclined to release theseresources when they are not in use. Moreover, because resourcesbecome available sooner, the cloud can support more users with fewer physical resources than would be needed in atraditional hosted environmenta direct cost savings forIBMs CIO Office.

As our internal success with cloud-based development and test workloads spread, so too did the demand. To date, 13,000 virtuamachines have been provisioned at five data center locations, anmore than 2,100 IBM users regularly tap into the developmentand test cloud.

Initially enabled only on a private cloud, our developmentand test activities are supported today by both private andpublic clouds. IBM SmartCloud Foundation provides thecloud infrastructure management software needed for privatedevelopment and test clouds. Weve also facilitated the develop-ment of select applications in an enterprise-class public cloudenvironment via IBM SmartCloud Enterprise and, withthe recent acquisition of SoftLayer, will be transitioning toSoftLayer in the near future. SoftLayers IaaS solution offersthe cost benefits of a multitenant, pay-as-you-go architecture without the security risks.

8/13/2019 CIW03082USEN

7/16


SCE has enabled the development of other cloud-based services,including a multiplatform mobile application developmentenvironment, which enables developers to expedite the designand deployment of mobile application solutions. A developerdesktop service provides developers with powerful secondary workstations for robust application development, testing, model-ing and simulations. These desktop environments provide ampleCPU power, memory and RAM, and faster I/O and data transfer

when developers personal laptops fall short.

IBM Development and Test Cloud benefits summary Resource provisioning reduced to as little as 1 hour from

5 days or longer

More efficient resource utilization through higher vir tualization

Increased availability of system resources as developers are morelikely to de-provision images sooner

Reduced labor spent building and suppor ting development and testenvironments and deploying associated middleware

Standard builds, reducing defects and the costs of customconfigurations

Analytics workloadsIn four years, business analytics have grown from a corporatedirective to an essential IBM service, largely due to the successof our internal analytics cloud, Blue Insight. The cloud-basedtechnology has literally changed the way the business operates,providing high-volume, real-time analytics and automatingbusiness decisions for nearly 300,000 employees today.

Blue Insight grew out of the need for a better way of generatingand delivering business intelligence (BI) insights. IBM washeavily invested in BI projects enterprisewide, but applications were being built separately and autonomously by individualbusiness units, creating silos of business intelligence. More than

100 different instances with multiproduct BI tools were beingused across the company, each with over 100 users, and runningacross more than 300 data sources. In addition, an estimated50 to 75 percent of new BI projects were not being co-located.

The lack of global data standards, systems and practices forIBMs BI projects resulted in massive duplication of effort andcostly maintenance and support. Conflicting content raised

questions about the integrity of metrics and data sources. And with the majority of BI applications deployed on individualdesktops, it became difficult to share BI content effectively. Withthe average BI application taking around six months to deploy,IBM estimated that each BI team and infrastructure requireda minimum six-figure investment. This reduced the overallnumber of BI initiatives that could be funded.

As many as 50 new and sizeable BI installations were on thedrawing board when Blue Insight was deployed. Blue Insightoffered a way to cull waste, duplication and inefficiencies by:

Standardizing IBMs BI tools, and consolidating andcentralizing current analytics capabilities to achieve greatereconomies of scale

Delivering analytics as a utility-like self-service at a cost that would make the capability universally accessible, whileeliminating capital and maintenance costs for individualbusiness units

Democratizing data and insights for the purpose of makingthe intelligence available across the enterprise, whilepreserving individual business units content ownership

Providing the infrastructure elasticity to expand IBMs

BI applications and users as rapidly as the business dictates.

8/13/2019 CIW03082USEN

8/16


Blue Insights power comes from its ability to draw insightsfrom hundreds of information warehouses and data storesacross IBM and generate analytics on more than a petabyte(one million gigabytes) of federated data. By turning raw datainto intelligence for our business communities, Blue Insightenables IBM to deliver more value internally and externally,more quickly and cost-effectively, as illustrated in Figure 2.

Figure 2. Centralized business intelligence delivered as a service. Blue Insightmakes high-quality business intelligence easily consumable, splicing togetherand analyzing critical data from hundreds of sources while enabling IBM usergroups to retain ownership of their BI applications.

Senior Executives

Sales andMarketing

ProductDevelopment

Fulllment

FinanceHR

Inquiries

Blue InsightBoarded analytics applications

Raw data

Intelligence

Hundreds of information warehouses

Blue Insight transforms IBM Treasury OperationsIBM Treasury Operations conducts business from multiplelocations around the world. Prior to Blue Insight, bank-relateddata was gathered and analyzed independently at eachlocation. There was no way for Treasury employees to accesscommon global data, and their ability to perform ad hocinquiries and reporting was limited.

Blue Insight enabled the deployment of Treasury Workstation,a single gateway for all banking communications via acommon data warehouse. With centralized global treasuryoperations and a uniform source for data, IBMs financialteams can create the standard and ad hoc reports needed forglobal data analysis. They have the global visibility to under-stand counterparty limits and exposures and to perform cash forecasting and treasury position analysistasks that weretoo complex and time-consuming to perform before.

IBM sales teams are using Blue Insight to gain a deeperunderstanding of each clients needs, not just for their ownIBM product group or region, but for all IBM products andservices worldwide. Our product development teams are usingBlue Insight to analyze sales information, industry trends andcustomer perceptions more efficiently and to adjust theirproduct planning and development specifications accordingly.Human resources is using Blue Insight to evaluate IBMs internaskills and capabilities against the emerging needs of our custom-ers, so we can identify and fill shortages. Finance is using BlueInsight to make revenue projections, evaluate risks and reduceour exposures. In all of these instances, data collection andanalysis tasks that formerly took weeks or longer to accomplishcan now be completed in hours or less.

8/13/2019 CIW03082USEN

9/16


With the deployment time for new BI applications down to two weeks, IBM business communities are extending the ways theyuse Blue Insight. They are thinking about new kinds of intelli-gence and leveraging Blue Insight to find it. For example, theyare tapping into social media to know their customers better toimprove their experience. And with over four million BI reportsbeing generated quarterly, they have come to depend on BlueInsight to get their jobs done. Today the analytics-based software

as a service (SaaS) capability is home to more than 500 boardedanalytics applications. Its subscription-based pricing modelenables participating IBM departments to stretch their BI bud-gets farther and cover more users. Managers can leverageanalytics to solve the days business problems and improveforecasting without worrying about affordability.

For IBM as a whole, the financial benefit continues to bedramatic. When polled, our top 20 Blue Insight consumersestimated the value of the analytics and intelligence theyvereceived to be at least USD $300 million. And those 20 repre-sent only a small percentage of the total number of Blue Insightusers at IBM. Blue Insight is also savings tens of millions ofdollars through its multitenant cloud infrastructure (includingmiddleware and IBM Cognos software); common operationalsupport and management, service definitions and boardingprocess; standard security; automated provisioning and adminis-tration; reduction in custom BI applications and displacement ofthird-party analytics software licenses. Business groups no longerfind the need to create their own BI environments to have theirrequirements met. In fact, since Blue Insights inception, fewerthan 20 alternate BI environments have been developed.

Finally, it is important to point out what Blue Insight means forthe future of IBMs intelligence capabilities. Current cost savingsand business value are just the beginning of a larger analyticstransformation at IBM. Many employees are just starting to tapinto Blue Insights potential, speculating about a critical piece ofintelligence theyd like to have and using Blue Insights enor-mous data resources and applications to locate it. As IBM users

become increasingly educated about the possibilities, they aremore likely to turn to Blue Insight for intelligence they onlydreamed about before.

We fully expect next-generation BI tools to have the abilityto leverage data in any form and deliver insights that are inter-preted and tailored for consumption by particular users, basedon their interests. These tools will be able to listen and act on

social media automatically, in real time, to identify businessopportunities and customer prospects. They will model out-comes and make decisions for users, tell them which plan to exe-cute or which customer to target. And the cloud will continue tomake these advances possible by enabling algorithmic processingon a monumental scale and orchestrating the analytics to deliverdesired insights rapidly and cost-effectively.

IBM Blue Insight benefits summary Standardized business intelligence and analytics capabilities,

delivered as a service

Expected savings in the tens of millions over five years

Business intelligence valued in the hundreds of millions so far

Elimination of duplicate applications and infrastructures for dif ferentanalytics requirements

Data and report ownership retained by par ticipating business units

Powerful, highly elastic infrastructure, enabling more rapid deploy-ment of BI capabilities

Ubiquitous user access to enterprise insights and intelligence

Usage-based subscription model, stretching BI budgets far ther and

covering more users Greater flexibility of the BI skilled workforce

8/13/2019 CIW03082USEN

10/16


Storage workloadsEven with IBMs 10 petabytes of operational storage, skyrocket-ing data volumes are necessitating increased storage capacity,better utilization of storage resources and automated tools tomanage it all. Like many of our clients, IBM is experiencing a25 percent annual growth rate in internal storage. Often thatrate spikes to 40 percent or more. While it is not possible to stopthe escalating demand for storage (due in large part to the rising

use of mobile devices, social media and rich media like voice and video), storage clouds make it possible to lower the cost per unit,or byte, of storage appreciably. These clouds can significantlyoffset the increased cost associated with soaring data volume.

For IBM, it became increasingly apparent that traditional stor-age solutions lacked the crucial scalability needed to serve ourlarge end-user communities. By moving internal storage capacityto the cloud, we have been able to virtualize storage resources toslow the demand for new storage and drive higher utilizationrates. We have also lowered labor costs through the automationof storage management functions.

Today 75 percent of IBMs internal file storage (about1 petabyte) is provided through a global file storage cloud. Builton a general parallel file system, this file storage cloud enablesemployees to share data, make backups and manage their storageneeds with self-service capabilities. This global file storage cloud was effectively IBMs first cloud, exuding the characteristics ofa cloud before the technology came to be defined. It provideselastic capacity, advanced virtualization and data replication forcost-effective business continuity and disaster recovery. It iscurrently accessed by more than 130,000 users and many ofIBMs internal applications.

IBMs block storage cloud builds on the success of our global filestorage cloud, leveraging similar storage automation functionsto address demand while containing costs. With block storagerepresenting 9 petabytes of IBMs internal storage inventory, it isour predominate class of operational storage. Understandably,cost-effective scalability is critical.

In 2010, we began work to transform IBMs Managed StorageServices, our on demand storage offering, into a block storagecloud with automated provisioning and storage tiering. TheIBM SmartCloud Virtual Storage Center grew out of that effort. Today it enables companies to use heterogeneous storageresources in a cloud-based, virtualized storage environment,optimizing storage capacity, provisioning and utilization. Itshypervisor capability facilitates virtualization management.

An intelligent storage service catalog streamlines storagerequests and provisioning. The intelligent information lifecyclemanagement (IntelligentILM) capabilities inherent in the hyper- visor automate the placement of data on appropriate drives inappropriate tiersin many cases, lower cost tiers. Using anadvanced block storage subsystem, IntelligentILM dynamicallyassigns each type of data to its optimal storage tier based onestablished cost/benefit profiles. It enables IBM to achieve thebest balance of storage cost and performance at any point intime and, in so doing, enables us to realize a much lowercost-per-byte of storage.

That lower rate is already a key factor in IBMs ability to handlethe exploding growth in storage demand, as well as the total cosassociated with that demand. In 2011, projections indicated thateven with 25 percent growth in storage volume each year, IBMscost for block storage should remain flat through 2015, wher-ever IntelligentILM-managed storage is used. Put another way,in 2015 IBM expects to pay the same total amount that we paidfor our storage inventory in 2011 while having the use ofapproximately 2.5 times more storage volume.

So far, where IBMs Virtual Storage Center has been imple-mented, the results have been striking in terms of savings, per-formance and deployment speed. At present, nearly all of IBMs virtualized storage at our major U.S. data centers leverage our Virtual Storage Center in the cloud. With that, the cost-per-byteof storage has dropped nearly 50 percent, enabling IBM tomitigate the increasing costs of much of the growth in storage volume in these locations.

8/13/2019 CIW03082USEN

11/16


Overall, IBMs cloud-based Virtual Storage Center is projectedto save 30 to 40 percent in storage infrastructure and manage-ment costslargely the result of end-to-end automation forstorage request, provisioning and tiering processes.

IBM Storage Cloud benefits summary Overall savings of 30 to 40 percent in storage infrastructure and

management costs Nearly 50 percent reduction in the cost-per-byte of block storage,

offsetting the r ising cost of volume growth

Inherent scalability to accommodate skyrocketing capacitydemands while controlling costs

Automated storage tiering, dynamically moving data to alternatetiers to optimize cost and performance

Better storage performance via elastic capacity, advancedvirtualization and data replication for cost-effective businesscontinuity and recovery

Collaboration and social business workloads

While IBM employees have used desktop-installed web confer-encing capabilities for years, moving this function to the cloudsimplified access, allowing our global workforce to collaboratesecurely regardless of device or location. Usage escalated drasti-cally. In the clouds first year, the number of meeting minutesnearly doubled.

Today IBM Connections, our cloud-based social networkingplatform for the enterprise, provides our employees with socialmedia capabilities adapted for workplace collaboration, allowingthem to engage with fellow experts around the globe in thecontext of their critical business processes. Externally, it enablesus to locate and stay connected with clients, build awareness ofour products and services, and innovate. Internally, it increasescollaboration and productivity by simplifying the import ofexternal information. It enables employees to conduct conversa-tions anywhere, anytime via mobile devices.

As of this writing, over 600,000 IBM users, partners and custom-ers have created profiles in IBM Connections, and they are col-laborating in large numbers. They have initiated 75,000 blogs,170,000 communities and 178,000 forums, and they have shared well over a million files. Over 500,000 of them are dispensingand storing information in 100,000 wikis, which total 1.6 millionpages and record nearly 475,000 page views each month.IBM employees are also using IBM Connections to follow

activity streams, tracking over 400,000 recent activities thatare occurring within the network. This enables employees tofollow the progress of specific projects and contribute to theirdevelopment by posting their own comments, documentsand links.

IBM Connections as problem-solver By bringing people together in new ways, IBM Connectionsprovides an enterprisewide platform for the generation of newideas and the dispersal of solutions to solve specific problems.One of these problems centered on locating IBM employeeswho had specif ic expertise. With over 400,000 employeesacross the globe, each with dif ferent skill proficiencies, it wasvery difficult to pinpoint individuals with proficiency in lesscommon areas.

The IBM Exper tise application was developed to solve thatproblem. It combines the power of IBM Connections withmultiple data capturing and search tools to rapidly locatepotential experts. The application leverages skill-relatedcontent supplied by employees in their Connections profiles,searching for specific keywords that reflect desired expertise.It finds employee matches in a matter of seconds, enabling

knowledge to be shared quickly. In the future, IBM expects touse employees Connections profiles to locate the specificassets that they have created, making those assets moreaccessible to other employees who can benefit from them.

8/13/2019 CIW03082USEN

12/16


Currently over 50 million minutes per month are providedthrough IBM SmartCloud for Social Business, which hasreduced costs and increased user productivity by providingcloud-based access to robust conferencing capabilities. IBM hasalso moved online meeting and event support to the cloud.IBM SmartCloud Engage provides cloud-based access to a variety of business tools and services that facilitate online events,including meeting invitations, notifications, real-time meeting

management and feedback surveys. The sharing feature allowstens of thousands of registered users to share select documents,applications and activitiesor their entire desktopwith fellowemployees or clients.

IBM Collaboration and Social BusinessCloud benefits summary Single application that enables employees to converse in whatever

style suits them bestblogs, forums, communities, etc.based onmutual interests and objectives

Dramatic increase in collaboration and productivity among

IBM employees

Faster development of innovative, new solutions, includingIBM Exper tise, which facilitates the location of subject matterexperts f rom among IBMs huge employee base

Shared source of information which can be searched and applied tosolve business problems

Better connectivity with clients and business par tners to buildawareness of and promote IBM products and ser vices

IBMs other successful workloadmigrationsIBMs IT organization is in various stages of migrating two othecommon workloads to cloud. Like the workloads discussed inthe previous section, desktop and production application work-loads have a strong affinity for the attributes that best align withcloud computing. As a result, they are proving to be a good fitfor cloud, with very positive results.

Desktop workloads

Desktop clouds centralize the administration and managementof end user desktops, facilitating the virtualization of desktopapplications and data using thin clients. IBM is in the earlystages of desktop cloud deployment, with thousands of produc-tion users accessing a private desktop cloud in IBMs ChinaDevelopment Lab and a considerable number of call centerusers accessing a desktop cloud in the U.S. and India.

While IBMs desktop clouds have simplified desktop administration and management and improved desktop security across theenterprise, IBM has learned that the key to deriving value fromdesktop cloud is choosing the right user population and pairingit with the right reference architecture. There is much to begained, for example, from a desktop cloud that supports callcenter personnel who provide very basic help desk servicesbecause it allows a simple desktop cloud implementation andavoids having to support and pay for desktop image persistence. A desktop cloud also allows for increased security at a lower cosso it is a good choice for user segments where unauthorizedaccess and other vulnerabilities are of strong concern.Conversely, a desktop cloud may not be a good choice for user

segments that require lots of custom software and device driversand whose thin clients are not highly standardized.

By understanding end users roles, usage patterns and businessrequirements, it is possible to segment users and to deploydesktop cloud where it will achieve the most success.

8/13/2019 CIW03082USEN

13/16


Production application workloadsIBMs production application cloud was implemented to reducethe cost of managing our internal application portfolio. Its initialimplementation was designed to provide low-cost infrastructureas a service (IaaS) capacity for IBMs non-critical applicationsapplications which have lesser service level requirements foravailability and support, and as such, are deemed low-risk.

Potential applications were identified for the initial productioncloud implementation, and the first of those have been migrated.Not unexpectedly, however, many of IBMs legacy applicationshave been found to be less suited for cloud. In some cases, theseproduction workloads have already been sufficiently optimizedfor their current environments, reducing the need for and bene-fit of optimization in the cloud. Others are supported by manu-ally administered controls and compliance checks that limit whatthe cloud can do to improve current production speeds andoutcomes.

Unlike newer applications that are born on the cloud and,so, specifically designed to run in a cloud environment, olderapplications are often not flexible enough to derive meaningfulbenefits from the cloud. They can take advantage of cloudsrapid provisioning, for example, but cannot bypass current man-ually administered controls that slow production considerably.For that reason, many of IBMs legacy applications are beingexamined in the context of the cloud environment to determine what it will take to bring the two together cost-effectively.Similarly, IBM is looking at ways to automate the manualprocesses that are inhibiting applications successful move tothe cloud.

Broadening cloud use across the IBM ecosystemIBMs Self Service Application Environment (SSAe) project ishelping cloud computing take a giant leap forward among theIBM user community. Cloud-based SSAe leverages IBMsextensive collection of application assets, standardizes themand makes them available to employees across the enterpriseat a predictable, affordable rate. In so doing, it encourages

IBMs lines of business to tap into the cloud to easily create,alter and retire applications.

Instead of expending the time and resources to develop newapplications, users can download desired applications fromthe clouds self-service catalog and get them up and runningwithout IT involvement. Support is available if it is needed, butno application development skills are required. SSAe providesthe tools to simplify the building and dissemination of avariety of applications. And because these applications arecomplementary to IBMs cloud-based information access frameworks like Blue Insight, users can benefit from theanalytics, innovative solutions and problem solving their fellow

employees are working on globally.

In the near-term, IBMs goal is to migrate as many non-criticalapplications as possible from their traditional hosting environ-ments to the production cloud, refining the platform as ourmaturity increases in this area. More critical production applica-tion workloads are expected to be migrated in the future,informed by our early experiences and the processes and toolsdeveloped as part of those experiences. The value of IBMsproduction application cloud will be realized primarily through

virtualized hosting and lower support costs enabled by increasedautomation and workload standardization.

8/13/2019 CIW03082USEN

14/16


We are also beginning to use new platform as a service (PaaS)technologies to cloud-enable production workloads. PaaStechnologies like IBM Workload Deployer (IWD) andIBM PureSystems have the potential to enable self-servicedeployment of certain production workloads in the cloud,significantly lowering costs. These technologies leverage user-defined, standard images and patterns to reduce setup time andspeed application deployment. The standardized images enable

automated processing, including resource provisioning whichcan significantly improve utilization. They also increase flexibil-ity, making it easier for application development teams to deploybug fixes and enhancements.

Clouds role in reinventing IBMs businessOver the last century, IBM has pursued a series of technologicalopportunities that have transformed our business and, in manycases, business at large. Cloud computing is proving to beanother of those opportunities.

Cloud computing is creating a sea change, not just in howIT resources and services are delivered, but in how the businessdoes business. And cloud has proven its potential to marrysubstantial cost savings, flexibility and efficiency to businessreinvention:

Helping IBM integrate information across the operation toimprove problem response and make smarter businessdecisions

Enabling work teams to tap into enterprise know-how,resources and applications that were formerly unavailable tothem economically, technically or geographically

Putting powerful analytics in the hands of sales, marketing,finance, HR and other IBM business groups to solve businesschallenges, improve efficiency and exploit opportunities forrevenue and growth

Helping researchers accomplish in hours what used to take weeks: complex data calculations and comparisons, modelingand simulations

Helping the business bring new and innovative solutions tomarket faster by facilitating experimentation, collaborationand social business.

This transformative experience has been seminal to the develop-

ment of the IBM SmartCloud portfolio of next-generation cloudtechnologies, services and solutions. The IBM SmartCloud port-folio leverages the same foundational technologies used in IBMown cloud implementations as well as client implementations,and it leverages what weve learned in those implementations,namely the importance of workload affinity and service choicein the deployment of cloud computing. It combines the costsavings of a shared cloud environment with the service optionsmore typical of a private environment: security, applications,service levels, management, support services and the like.For the first time, companies are able to tailor their cloudenvironments to match the requirements of their workloads.

Finally, its important to point out that IBMs ongoing success with cloud computing has a lot to do with our culture.Organizations need to be willing to modify or even jettisontheir existing processes if they are to take advantage of newtechnologies like cloud. IBM has always thrived on technologicachange. Our IT organization is often the first testing groundand proving groundfor new technologies, and cloud is nodifferent. Today cloud is continuing to make its way across theIBM ecosystem. It has already had a profound impact on theIBM enterprise, but in our view, the transformation has only just begun.

8/13/2019 CIW03082USEN

15/16


ConclusionCloud computing has significantly altered the IT landscape atIBM, winning over early adopters with improved productivityand performance, and far exceeding initial expectations forusage. Its magnetic appeal speaks to the speed, flexibility andself-service capabilities of the cloud, especially for IBM usersengaged in development and testing, storage and analytics activi-ties. These user groups can really appreciate clouds attributes inthe context of their workloads unique demands.

Clearly, the value that cloud brings to these workloadsand tothe organization as a wholewill increase with the maturity ofthe model but also with the experience of those business unitsthat take advantage of it. IBM recognizes cloud as a majorenabler of ITs evolution and the future of IT service delivery. Towards that end, we are actively expanding the depth andcapabilities of our cloud portfolio to help clients uncover newsources of revenue and to capitalize on the potential for businesstransformation, but more fundamentally, to provide them withthe necessary strategic guidance and tools to make the rightcloud choices. As a part of that, we are also focusing on thedevelopment of applications specifically designed for cloud thatbest exploit the services, capabilities and considerations that areinherent in the cloud platform on which they run.

Understanding each workloads affinity for the cloud is essential,and it should be central to any cloud adoption discussion. IBMsstructured methodology for workload analysis is fueled by this

belief and derived from our cloud experiences internally and with clients. Today that methodology is helping companiesanalyze and prioritize application workloads for cloud, factoringin potential cost and migration impacts. After all, choosingthe right workloads can deliver extraordinary economic andoperational value.

For more information To learn how IBM is helping organizations migrate the right workloads to the cloud, please contact your IBM representativeor IBM Business Partner, or visit the following website:ibm.com /smartcloud

You can also follow us on Twitter at:www.twitter.com/ibmcloud and on our cloud computing blog at: www.thoughtsoncloud.com

About the authors Michael Sylvia Director, Advanced Technologies

IBM Office of the CIODistinguished Engineer, IBM Academy of Technology

Brian Peterson Senior Technical Staff Member, Advanced TechnologiesIBM Office of the CIO

Steven Uniack Executive Architect, Advanced TechnologiesIBM Office of the CIO
http://www.ibm.com/smartcloudhttp://www.ibm.com/smartcloudhttp://www.twitter.com/ibmcloudhttp://www.thoughtsoncloud.com/http://www.thoughtsoncloud.com/http://www.twitter.com/ibmcloudhttp://www.ibm.com/smartcloud

8/13/2019 CIW03082USEN

16/16

Copyright IBM Corporation 2013

IBM Global ServicesRoute 100Somers, NY 10589U.S.A.

Produced in the United States of America July 2013

IBM, the IBM logo, ibm.com, IBM SmartCloud, Cognos, and PureSystemsare trademarks of International Business Machines Corporation in theUnited States, other countries or both. If these and other IBM trademarked

terms are marked on their first occurrence in this information with atrademark symbol ( or ), these symbols indicate U.S. registered orcommon law trademarks owned by IBM at the time this information was published. Such trademarks may also be registered or common lawtrademarks in other countries. Other product, company or servicenames may be trademarks or service marks of others. A current list ofIBM trademarks is available on the web at Copyright and trademarkinformation at ibm.com /legal/copytrade.shtml

This document is current as of the initial date of publication and may bechanged by IBM at any time. Not all offerings are available in every countryin which IBM operates.

The performance data discussed herein is presented as derived under specificoperating conditions. Actual results may vary. It is the users responsibility toevaluate and verify the operation of any other products or programs withIBM products and programs.

THE INFORMATION IN THIS DOCUMENT IS PROVIDEDAS IS WITHOUT ANY WARRANTY, EXPRESS ORIMPLIED, INCLUDING WITHOUT ANY WARRANTIESOF MERCHANTABILITY, FITNESS FOR A PARTICULARPURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the termsand conditions of the agreements under which they are provided.

Statements regarding IBMs future direction and intent are subject to changeor withdrawal without notice, and represent goals and objectives only. Actuaavailable storage capacity may be reported for both uncompressed andcompressed data and will vary and may be less than stated.

Please Recycle

CIW03082-USEN-03
http://www.ibm.com/legal/copytrade.shtmlhttp://www.ibm.com/legal/copytrade.shtmlhttp://www.ibm.com/legal/copytrade.shtml

Documents

CIW03082USEN