Cloudsourced

Tony Kerrison, CTO of ING, on How Cloud Computing is Changing IT

Cloud computing pioneer Kerrison, who is building a private cloud and investigating the use of public clouds at the Amsterdam banking giant, believes cloud computing will profoundly affect IT executives' jobs.

By Penny Crosman

When last I spoke with Tony Kerrison, who today is CTO at Amsterdam-based banking giant ING ($1.6 trillion in assets), he was global head of IT at Merrill Lynch. Even way back in 2008, he was building a cloud-like infrastructure for the Thundering Herd, creating a "stateless infrastructure" for data center servers and rolling out virtual workstations. Three years later, he is implementing his cloud-computing vision at ING and developing cloud computing standards within the TM Forum's Enterprise Cloud Leadership Council. He spent time with us this week sharing information about the cloud computing projects he's started at ING, how cloud computing is changing the working lives of bank technology leaders and how the industry can and should be creating standards and requirements for cloud computing providers.

Bank Systems & Technology: To what extent has ING adopted cloud computing?
Kerrison: For the last year, we've been building our own private cloud, which is helping us understand how a cloud infrastructure can work in our environment and how to apply the resources to support it and put more and more applications in our cloud environment.
At same time, although we don't use any public cloud infrastructure at this point, we're watching it carefully, in particular utility applications such as email-as-a-service. We're also looking at how we can develop cloud testing services, because we've built an internal center of excellence around testing.
BS&T: What is the scope of your private cloud, and what types of applications are you running in it?
Kerrison: The infrastructure we've created is x86 based and we're using self-provisioning tools on top of that. We're restricted from talking about specific vendors. Our application groups are able to spin up their own environments for testing and development in the cloud. We've also built a grid on the same infrastructure and we're deploying some of our commercial banking applications there. We're being cautious, we're focused on low priority and low criticality apps at this stage, but the results so far indicate we can push ahead.
BS&T: Can you share what percent of your applications today are running on some kind of cloud, and where you see that one to five years from now?
Kerrison: Somewhere around 5% of our application portfolio today are on our private cloud. We expect that to grow significantly in 2011 and we could be up to around 20% at the end of the year. We're going to learn as we go forward and reexamine the complexity of the environment we have, which apps it makes sense to cloud-enable, and how do we invest the time and money to make them cloud enabled. This will definitely be a year of deployment -- in 2010 we built the foundation to deploy in 2011.
We also spend a fair amount of time implementing virtual desktop infrastructure at ING that could evolve into an external cloud service, which might be called desktop as a service. All our remote access comes through our VDI environment, as well as about 10% of the population of ING. That's growing exponentially.
BS&T: When you say remote access, do you mean people who work from home or in remote offices?
Kerrison: It's both, actually. It gives ING more mobility, it opens lots of opportunities for us. People are able to do all their daily work from another location, such as an office on an ING campus or from home. It's a very good, flexible way for people to do their work.
BS&T: Do you see rolling VDI out throughout the bank, with eventually all desktops being virtual?
Kerrison: We offer four different types of desktops. There are still laptop users who need to work on planes, trains, and such. There are some high-power users who don't use virtual desktops. But the majority of ING users will be on VDI base as we head into 2012, we think that will just grow over time.
We do VDI not only on standardized thin clients at offices and home, but we're also rolling it out through mobile devices such as the iPad.
BS&T: Does ING have a desktop instance or apps that run on the iPad already?
Kerrison: We've got some applications, such as email, in test right now.
BS&T: What would you say are the main benefits for banks of VDI; is it the flexibility, is it less expensive?
Kerrison: It's a number of dimensions. The underlying infrastructure will be less expensive, although the cost of PCs and laptops is coming down. But over time, you get the flexibility and independence that goes with the ability to have any device connected as the desktop. It also drives down real estate costs as people can work from home or desk sharing.
BS&T: Does that parallel some of the benefits of your private cloud?
Kerrison: The private cloud helps us optimize the equipment in the back end; virtualization and consolidation provide significant benefits for us. The provisioning capability and the standards we're setting will create more productivity for the application teams, starting with testing and into production.
BS&T: At what point would you say you crossed the line from large-scale virtualization to cloud?
Kerrison: Over the last couple of years, we've been pushing hard on virtualization. In 2010, we began further improving our use of server virtualization for productivity and efficiency. As we set standards around self provisioning and educated people on what virtualization and cloud services were, it became clear we were moving toward a cloud environment rather than straight virtualization. I don't think one day we woke up and said today it's virtualization, tomorrow it's cloud, but the things we were doing were leading us to become a service-based IT organization, which moved us toward a cloud infrastructure. Over time, although we've currently focused on our private cloud, the goal is to look for more opportunity to migrate certain utility applications from the private cloud through a hybrid environment to the public cloud. As the regulations, standards and service levels become clear, we should be trying to exploit those low-cost environments.
BS&T: Some bankers feel that they'll never put their most sensitive data or highest speed apps on a cloud. Where do you stand on that? Do you feel this is where IT is heading and it's only a matter of time, or are there any apps you would never put on a public cloud?
Kerrison: I would say, never say never. The reality is all of us have to manage a portfolio of different applications that have different attributes and requirements. Once you determine the technical capabilities and requirements of your apps, you then need to understand the regulatory requirements with regard to where data is being stored and how much control you have over that data.
Our job in running technology in banks in the future will be all about managing the portfolio. I see my role as managing a portfolio of apps across four segments: our legacy environment, a high availability infrastructure that's not even potentially cloud-like; a private cloud infrastructure that provides flexibility, efficiency, scalability and control, managed in such a way that we can completely understand the implications of the data; hybrid computing environments (that use a mix of internal and external cloud services); and public clouds. IT's role fundamentally changes to become one of figuring out where the optimal place is to run an application with a specific set of data, and I think even a lot of traditional outsourcing dissolves. As you move closer to the public cloud, you're getting a lower cost point. Our job is to balance all those aspects of running that portfolio to optimize it as much as we can. Even now, you can see tools and software technology emerging that's enabling us to do that analysis, to think through the different services out there.
BS&T: Are you referring to tools like business service management software that can show you the performance of each alternative -- this one has a five second response time, that one has a five microsecond response time?
Kerrison: Exactly. Our job is to broker what's available in the marketplace to be able to use the optimal service solution. It's no different than comparing a Salesforce.com to apps you run yourself. Our job is to weave all these solutions together. If you don't want to use 150 of these services, you may end up going to larger service providers.
BS&T: How has being in the TM Forum's Enterprise Cloud Leadership Council helped you define your strategy and determine where you're going to experiment and move ahead with cloud?
Kerrison: First, the ECLC is a community of people who are in the same situation I'm in, trying to evaluate these opportunities in the same way, so it's a great forum to work collaboratively. Second, the group is trying to add more clarity to cloud computing by setting standards and creating real life examples of how to make these services work in a way that benefits everyone. The other thing that's important about ECLC is that we all contribute to it. For example, within ING we're contributing information about processes for cloud environments, for instance, what's the cloud version of ITIL? We have some IT staff who are very expert at ITIL and we're using that expertise to help us along. Another thing we've worked on and are bringing to ECLC is training, we've worked with all our major technology providers to come up with a training course that we can offer to ING staff and as a foundation for members of ECLC to make that standardized training.
The establishment of standards is so critical to being able to evolve cloud services -- without standards we'll never be able to exploit the marketplace in the right way.
BS&T: Is it possible to force the vendors to make their offerings interoperable, so you're not stuck on one cloud, unable to work with any other?
Kerrison: Absolutely. In developing the training, for one example, we've had collaboration from IBM, HP, Cisco, EMC, Microsoft, vmWare, Virtual Clarity and other vendors to help develop the courses and the book that supports the training.
BS&T: It seems critical for competitors to work together, otherwise what vendors call "cloud" or "open" is really not.
Kerrison: I agree. It's not easy to get vendors to work together, most have their own IP they're trying to protect. But we've found that it's in everybody's interest to drive to these standards, and after a while make sure people are working together. This is also true within ECLC -- there are competitors of ours that are also members, but we don't let that get in the way because ultimately what we're trying to do is drive toward the bigger picture.
When you think about it, the future of the cloud marketplace relies on collaboration, which typically isn't natural for vendors. I see the role of ECLC to help them see the benefits they can gain through creating these standards.

United Business Media LLC

Cloudsourcing vs Outsourcing

Cloud - the third generation of outsourcing

Srinivasan Sundara Rajan

Second Generation Outsourcing
Outsourcing is more and more viewed by the enterprises as a transformational activity, where by

Enterprises Utilizes Out Sourcing to enhance or refine the business capability
The Vendor/Supplier of out sourcing is viewed as a partner in transformation
It is more about inter dependence than dependence with a single vendor

Second-generation outsourcing is designed to facilitate customer intimacy and product leadership for an organization. This occurs through a transformation of IT architectures; alignment of IT with the business; and the optimization of business processes to take advantage of the new capabilities provided. These in turn lead to increased shareholder value. Second-generation outsourcing shifts the relationship between outsourcer and outsourcee from trusted supplier to vital business partner.
Cloud SourcingCloudsourcing refers to sourcing complete solutions to run your business from the public cloud. The solution provider that offers Cloudsourcing products or services is called "Cloudsourcer". The Cloudsourcer typically provides solutions that knits together cloud applications, cloud platforms and cloud infrastructure.
Cloud Sourcing Bettering Out Sourcing
The below table gives a list of areas where the cloud sourcing significantly out weights benefits of second generation out sourcing, which the enterprises can consider for their future business goals.

Category	Second Generation Out Sourcing	Cloud Sourcing	Comments
Upgrades , Patches and maintenance of existing software stack	POOR	GOOD	In traditional out sourcing, data centers exist inside premise and take significant time in upgrading larger work like database upgrades or operating system upgrades. These even disrupt the business capability needs
Quick Procurement of Services	FAIR	GOOD	The multi vendor model of the second generation out sourcing also brought in challenges for consistent discovery, request and provisioning of services, however the basic Cloud tenant enables a catalogue of services.
Retention of Key Subject Matter Expertise	FAIR	GOOD	Most out sourcing models retain the key IT staff with the Subject Matter expertise, however as the work is shifted to the vendor data centers this retention of key employees has been affected, Cloud sourcing ensures that the key staff is available for the interface with cloud provider.
Vendor Lock In Avoidance	FAIR	GOOD	Most of the projects are built using the Vendor proprietary features and hence the out sourcing model had to continue with vendor lock in, SaaS cloud sourcing makes the vendor product mapping irrelevant and gives flexibility to the enterprise
Legal and Data Proximity Compliance	GOOD	POOR	In out sourcing model the organization has complete control of data and cloud sourcing at this time is dependent on cloud provider's data centers.
Legacy Modernization, Consolidation of Redundant applications	POOR	FAIR	In the out sourcing model, merging existing systems or de-commission redundant systems is possible but it is mostly a manual and time consuming process, resulting in them getting postponed many a times. However Cloud Sourcing makes the physical blue print of the application transparent and hence redundant applications automatically cost less to the organization as their usage is likely to be less.
Federal, state and international legal compliance	FAIR	GOOD	In the out sourcing model, compliance projects are generally a urgent patch work to get it done before the dead line expires, In the SaaS model the product Vendors ensures compliance.
Thought Leadership in the Domain	FAIR	GOOD	Out sourcing model while supports the thought leadership from the Supplier, but it is limited to current business process flow of the out sourcing enterprise. However in Cloud Sourcing the multi tenant design ensures that best of thought leadership from various players in the particular domain is provided through a SaaS platform.
Minor Enhancements, ability to do quick fixes of the applications	GOOD	POOR	In the out sourcing model the contracts can be done in such a way to do a fixed capacity of work for minor enhancements, SaaS software is generally static over a period of time and depends on upgrades from the cloud provider

As per the over all analysis, Cloud Sourcing provides more benefits that traditional second generation that is in place and is worth considering for the enterprises.

via http://cloudcomputing.sys-con.com/node/1611752

CSC Launches CSC BizCloud

• CSC BizCloud, the only on-premise private cloud billed as a service currently available. Ready for workloads in 10 weeks

• BizCloud features CloudCompute, CSC's new infrastructure as a service (IaaS) architecture which is deployed in the CSC Trusted Cloud Datacenters.

• The CloudCompute infrastructure is built on the Vblock Infrastructure Platform from VCE, The Virtual Computing Environment Company.

• Vblock integrates the leading virtualization software, networking, security, computing, storage and management technologies from industry leaders Cisco, EMC and VMware. CSC's partnership with VCE enables the economic model that is unique to BizCloud.

• CloudCompute is offered in three packages: Standard, Enterprise and Enterprise Plus.

CloudCompute enables organizations to align workload requirements with the right CloudCompute package and service level.

Zimory enables T-Systems' IaaS cloud

Analyst: William Fellows
Date: 3 Feb 2011
Email This Report:451 Report Folder: File report »» View my folder »»

German IT services and systems integrator T-Systems is to offer a cloud IaaS service to its customers as part of its Dynamic Services for Infrastructure portfolio using cloud-enablement software from Zimory. While T-Systems had been noodling with Zimory technology for a couple of years since it was released in 2008, Zimory subsequently spent nine months as part of an RFP process and beat out competition from other startups, as well as bigger players such as The VCE Co and IBM to win the five-year deal. A report on the T-Systems service will follow.

The 451 take

With its recent funding, a new CEO and now an anchor customer for its cloud-enablement software, it looks as if Zimory has finally achieved liftoff. That its customer is T-Systems doesn't come as a great surprise, but neither should it be understated. Zimory beat out large vendors, and the opportunity could scale. T-Systems, which addresses the (mostly German) enterprise market for IT services, is owned by Deutsche Telekom, which, in turn, addresses the rest of the IT services market opportunity in Germany. As a customer (and the parent) of T-Systems, Deutsche Telekom will in the future resell the Zimory-based service to its target market.

Business

The T-Systems win comes after the arrival late last year of telco veteran Ruediger Baumann as Zimory's new CEO and a €4m ($5.4m) series A funding round in the fall, bringing total investment to €6m. The financing was led by Creathor Venture, High-Tech Gruenderfonds and VC Fonds Technologie Berlin (the vehicle of the IBB Berlin bank), T-Venture and KfW Bankengruppe (the bank of the German government). Deutsche Telekom, which provided some of the seed funding for Zimory – a 2007 spinout from the German telco – also participated in this round but has reduced its holding to 20%. Zimory has 22 staff.

The deal with T-Systems is exclusive in that for the price it has secured the Zimory software, T-Systems has agreed not to use another supplier for its cloud-enablement requirements for five years. It's paying an undisclosed up-front payment plus an annual charge based on the size of server farms with which T-Systems is supporting the service.

Outside of T-Systems, Zimory claims to be working in some government cloud deployments, closing in on German enterprise deals and in discussion with other service provider and hardware partners.

T-Systems IaaS cloud

The deal will see T-Systems make IaaS services available to its network customers initially. This is very much a private cloud offering in the sense that it is open only to customers that use T-Systems' network services. It's not available on the Internet or with a credit card. Moreover, while T-Systems' datacenters do provide a basic multi-tenanted resource to IaaS customers, they will be able to request isolated resources or resources shared with specific partners. Initially, the service will be VMware only, and while Zimory already has a Xen version of its platform operating in test mode, it doesn't expect T-Systems to support other hypervisors this year – although T-Systems does plan to support Microsoft, KVM, Xen and others over time. Zimory's technology manages T-Systems' server farms, provides self- and auto-provisioning of virtual machines and storage services, and claims to enable PaaS through template management. The service will be run from datacenters in Frankfurt, Munich and Berlin initially, but the service will extend globally over time.

The service will be priced according to VM size – 512MB to 64GB RAM and from one to 16 performance (compute) units where one unit is the equivalent of 10,000 TPMC. There's no up-front commitment, but there is a reservation system, and customers are billed per hour. There's an ongoing discussion about annual up-front commitments by customers and a resource commitment from T-Systems. T-Systems will provide packaged OS instances and other templates, although users can bring their own, too.

Zimory says the SLA will be equivalent to other IaaS offerings in the market, initially sold as a 'Silver Cloud' guarantee, presumably with Gold and Bronze offerings to follow. Customers of T-Systems' existing VMware-based Dynamic Services (aka cloud) offerings will be offered access to the service, although workloads won't migrate seamlessly.

T-Systems' existing Dynamic Services is the delivery point for its on-demand, Web 2.0, cloud and SaaS services. T-Systems hosts or manages end users' SAP, Exchange, Lotus, archiving, Microsoft Dynamics Nav or desktop requirements (or does it in a mixed mode). The bulk of Dynamic Services is virtualized with VMware, although it also runs Hyper-V. Besides running its own IT on Dynamic Services, T-Systems has also been moving parent Deutsche Telekom's infrastructure to the platform.

Context

In 2008, the Zimory Marketplace was introduced as a mechanism to aggregate spare capacity from supplier members and make it available on demand to marketplace participants in a brokered and federated model. Although two of T-Systems' datacenters (and, indeed, its customer base) were always seen as a target test ground for this, the Marketplace was never established. At one time, Zimory also counted Deutsche Post as a user and Fraunhofer ITWM, which built a hosted VM service for its developer community using Zimory's Xen implementation. At the time of the announcement, Amazon said it would participate in the marketplace if there were an opportunity, and Microsoft was referred to as a partner.

More recently, Zimory has rebranded its services, which are now packaged and marketed as a cloud-enablement service for enterprises and service providers. A single instance of what's called Connect is required in each datacenter as a local management hub and includes the core functions, enabling users to connect to local resources. A tool called Host is required to connect open source hypervisors to the cloud. Manage provides a manual management point for cloud users and an API. Multiple Managers can be connected to one or more Cloud Connector. Zimory Manage can also be connected to a single Connect instance to customize services for specific user groups. The Spree database hypervisor is now called Zimory Scale and will be (re)introduced in June. Its design point is to do for databases what virtualization does for operating systems, as a smart fabric in front of the relational database management system.

Competition

IBM (CloudBurst), HP (CloudSystem), Fujitsu, Unisys, The VCE Co, VMware, Microsoft, CA Technologies, BMC and Tibco are the competition on the vendor side. In the IT service sector (and excluding the above) are Accenture, CSC, Lockheed Martin, Capgemini, NTT Data, SAIC, Atos Origin-SIS, BT Global Services and Logica. Zimory's peers include Abiquo, Appistry, C12G Labs, CloudBroker, Cloudscale, CloudSigma, CloudSwitch, Cpanel, DediPower, DynamicOps, ElasticHosts, Elastra, Enomaly, Eucalyptus, Flexiant, fluid Operations, Hexagrid Computing, Inkspot, Joyent, Layered Technologies, Maatg, Morph Labs, Nimbula, OnApp, OpenStack, Platform Computing, ThinkGrid, Univa and Yunteq. Cloud.com arguably has the most marketplace traction in this segment.

Early adopter spells out Microsoft Azure's strengths and shortcomings

Microsoft Windows Azure customer builds 'technology that would not exist without the cloud'

By Jon Brodkin

IN DEPTH:

With Microsoft celebrating the first birthday of Windows Azure, few people are as well equipped to discuss the cloud platform's strengths and limitations as Joannes Vermorel.

Vermorel, the founder and CEO of a software development company in Paris, France, called Lokad, jumped on the Azure bandwagon well before the service even went online on Feb. 1, 2010. Because Lokad decided to move most of its infrastructure from hosted servers to the Microsoft cloud, Vermorel and colleagues had to rewrite the company's application infrastructure, and began doing so at the beginning of 2009, when Azure was in beta.

"It took more than 12 months to completely rewrite Lokad from scratch to have cloud-based technology," Vermorel said in an interview with Network World. "At the present time, we have a technology that would not exist without the cloud."

Vermorel, who blogs about his experiences with cloud computing, chose Azure over Amazon's Elastic Compute Cloud, and says it was the right decision. "Microsoft was the first and only one to have a clear vision of the cloud that includes the tooling experience," he says.

Still, there are a few shortcomings Vermorel urges Microsoft to work on. Redmond recently updated the Windows Azure user interface to make it more attractive and intuitive, but the service console still lags behind those offered by competing cloud services, Vermorel says.

"It was abysmal. Now it's OK. So the trend is good," he says.

Specifically, the user portal lacked multiuser support, which it now has, and is still slow despite a significant speed boost, he says. Before the update, it was tedious to deploy and redeploy services, and "was not very task-oriented," Vermorel says. "It's a more usable interface but they can still do much more and much better."

Lokad's system analyzes sales and demand figures, giving customers the forecasts they need to manage supply and demand, to optimize inventory and staffing levels. For retail, this can mean figuring out which products to keep in the warehouses and at the stores. For banks, it could mean keeping ATMs and branch offices stocked with the right amount of cash.

Lokad, founded in 2008 with an emphasis on data mining and grid computing, has just 11 employees, so managing a large data center would distract from its core mission. At first, the company rented servers from a hosting provider, but "most of the time those servers were doing nothing." At the same time, the hosters didn't provide enough scalability, so "there was never enough processing power" when demand for Lokad's system was high, Vermorel says. Lokad had just half a dozen servers in the U.S. and Europe, and setting up a new server required a big fee.

This hosting model was already a big improvement over owning and managing machines directly, Vermorel says, but what Lokad really needed was the flexibility offered by cloud services. Lokad's system requires heavy allocation and de-allocation of computing resources. If Lokad needs a few thousand servers to run computations for an hour, Azure can provide that. Microsoft and Lokad have developed a tight partnership, with Microsoft naming Lokad its Windows Azure Platform Partner of the Year in 2010.

Vermorel, who is teaching a course on cloud computing at a university, notes that Amazon EC2 has some advantages over Microsoft's Windows Azure. If you have lots of legacy code, and perhaps multiple operating systems including Windows, Unix and Linux, Amazon is the better option for porting applications to the cloud, he says. Amazon is also well suited for hybrid environments involving some in-house resources and some cloud-based resources, he says.

Amazon EC2 is "the most versatile cloud solution," but "probably the least productive," he says.

Azure, even Microsoft notes, is not very well suited for moving existing applications to the cloud. Although Azure has begun offering virtual machine hosting, in general it provides less control over the underlying computing resources.

Azure abstracts away a lot of the complexities, but required companies like Lokad to start from scratch. But that's all part of the natural transition to cloud computing, Vermorel suggests. Applications have to be completely re-engineered to take advantage of the unique benefits offered by cloud computing, including the ability to scale and quickly ramp up performance.

"It won't work out of the box," he says, noting that you also have to write reliability and availability features into the application and optimize it for security in a world that is much different than the classic client-server setup.

Putting an app on the Web allows remote access from nearly every conceivable point, and that means "every single component has to be strong. If one of your components is compromised, the others shouldn't fall down."

One complaint Vermorel has about Windows Azure is the hourly-based billing. If Lokad can improve the efficiency of its own process, cutting a task down from one hour to 50 minutes, Microsoft will still charge the same price. Microsoft should implement per-minute charges to allow greater pricing flexibility, Vermorel says.

But there are other advantages. "Unlike classical Microsoft products," bugs on Azure are typically fixed within two months, Vermorel says. Lokad is using numerous Microsoft products in addition to Windows Azure, including C#, the .Net Framework, and SQL Azure.

Microsoft is taking customer feedback seriously and expects to keep improving Windows Azure, says Amy Barzdukas, general manager of server and tools marketing for Microsoft.

"I think that generally speaking we are very serious about taking customer feedback and continuing to iterate," she says. "We are constantly looking at the pricing models just as we are with the feature sets."

Microsoft has released numerous improvements over the first year, including the ability to host Windows Server instances, and high-performance computing capabilities.

To promote the 1-year anniversary of Azure, Microsoft is issuing a press release touting customers T-Mobile USA, Travelocity.com and Xerox. T-Mobile used Azure to create a mobile software application called Family Room, while Travelocity opted for Azure for an analysis tracking system that helps improve its website's functionality and capacity, and Xerox build its Cloud Print service on top of Azure.

"What we're seeing with these customers, is they're turning to Microsoft for speed, ease and familiarity," Barzdukas says.

http://www.networkworld.com/news/2011/020111-azure-vermorel.html