cloud benefits and challenges/problems, business and operational...

Cloud Benefits and Challenges/Problems, Business and Operational Models, Cloud Computing

Cloud Strategy Partners, LLC

Sponsored by: IEEE Educational Activities and IEEE Cloud Computing

Course Presenter’s Biography

IEEE eLearning Library Cloud Ben. & Chal./problems, Bus. & Op. models, Cloud Comp. Econ. Transcript pg. 2 / 24

This IEEE Cloud Computing tutorial has been developed by Cloud Strategy Partners, LLC. Cloud Strategy Partners, LLC is an expert consultancy firm that specializes in Technology and Strategy relating to Cloud Computing.

Course Summary


This tutorial will go into more detail regarding Cloud Computing properties and benefits. We will look in detail at the elements such as “scalability and elasticity”, “availability and reliability”, “manageability and interoperability”, and “performance and optimization”. When properly utilized, Cloud Computing can provide these capabilities to the applications deployed on the Cloud. The requirements for such “proper utilization” will also be discussed. Even with such positive and interesting benefits, there are pundits who believe there are still fundamental problems in using Cloud Computing. In the analysis process of bringing applications to clouds, or in the transformation of a datacenter to Cloud, or along with practically any project considering the use of Cloud Computing, these concerns emerge and become impediments to Cloud implementation. In some cases, these are real concerns and in many cases they are not. These will be reviewed. A more detailed review of the business and operational models of Cloud Computing will be considered. When these are treated in sufficient detail, one can begin to think about “Cloud Computing Economics”, applying the Cloud model to different Cloud “service models” and different “deployment models”.

Transcript


Course Outline This Lesson will go into more detail regarding Cloud Computing properties and benefits. We will look in detail at the elements such as “scalability and elasticity”, “availability and reliability”, “manageability and interoperability”, and “performance and optimization”. When properly utilized, Cloud Computing can provide these capabilities to the applications deployed on the Cloud. The requirements for such “proper utilization” will also be discussed. Even with such positive and interesting benefits, there are pundits who believe there are still fundamental problems in using Cloud Computing. In the analysis process of bringing applications to clouds, or in the transformation of a datacenter to Cloud, or along with practically any project considering the use of Cloud Computing, these concerns emerge and become impediments to Cloud implementation. In some cases, these are real concerns and in many cases they are not. These will be reviewed. A more detailed review of the business and operational models of Cloud Computing will be considered. When these are treated in sufficient detail, one can begin to think about “Cloud Computing Economics”, applying the Cloud model to different Cloud “service models” and different “deployment models”.

One thing Cloud Computing does, beyond the improvements in the areas outlined above, is that Cloud Computing acts like a “IT Innovation Factor”, helping new applications come into being, which never would have been able to be implemented if it was not for Cloud Computing.

REF: Cloud Computing Definition: Components Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. This cloud model is composed of five essential characteristics, three service models, and four deployment models.

Essential Characteristics: On-demand self-service. A consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with each service provider.

Transcript


Broad network access. Capabilities are available over the network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, tablets, laptops, and workstations). Resource pooling. The provider’s computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand. There is a sense of location independence in that the customer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or datacenter). Examples of resources include storage, processing, memory, and network bandwidth.

Rapid elasticity. Capabilities can be elastically provisioned and released, in some cases automatically, to scale rapidly outward and inward commensurate with demand. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be appropriated in any quantity at any time. Measured service. Cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). This should allow services provisioning on a pay-per-use or charge-per-use basis. Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized service.

Service Models Software as a Service (SaaS). The capability provided to the consumer is to use the provider’s applications running on a cloud infrastructure (see below). The applications are accessible from various client devices through either a thin client interface, such as a web browser (e.g., web-based email), or a program interface. The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings. Platform as a Service (PaaS). The capability provided to the consumer is to deploy onto the cloud infrastructure the consumer-created or acquired applications created using programming languages, libraries, services, and tools supported by the provider. The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, or storage, but has control over the deployed applications and possibly configuration settings for the application-hosting environment.

Transcript


Infrastructure as a Service (IaaS). The capability provided to the consumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, and deployed applications; and possibly limited control of select networking components (e.g., host firewalls).A cloud infrastructure is the collection of hardware and software that enables the five essential characteristics of cloud computing. The cloud infrastructure can be viewed as containing both a physical layer and an abstraction layer. The physical layer consists of the hardware resources that are necessary to support the cloud services being provided, and typically includes server, storage and network components. The abstraction layer consists of the software deployed across the physical layer, which manifests the essential cloud characteristics. Conceptually the abstraction layer sits above the physical layer.

Deployment Models: Basic models defined in the NIST standard Private cloud. The cloud infrastructure is provisioned for exclusive use by a single organization comprising multiple consumers (e.g., business units). It may be owned, managed, and operated by the organization, a third party, or some combination of them, and it may exist on or off premises. Community cloud. The cloud infrastructure is provisioned for exclusive use by a specific community of consumers from organizations that have shared concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be owned, managed, and operated by one or more of the organizations in the community, a third party, or some combination of them, and it may exist on or off premises. Public cloud. The cloud infrastructure is provisioned for open use by the general public. It may be owned, managed, and operated by a business, academic, or government organization, or some combination of them. It exists on the premises of the cloud provider. Hybrid cloud. The cloud infrastructure is a composition of two or more distinct cloud infrastructures (private, community, or public) that remain unique entities, but are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting for load balancing between clouds).

Transcript


New emerging deployment models Federated cloud. The cloud infrastructure that involves multiple heterogeneous clouds from different providers that use a federation mechanism to share, access and control combined infrastructure and services. Federated cloud typically combines multiple private clouds and may include also private cloud. Federation members remain independent however having common policy in resources sharing and access control, including federated identity management. Cloud federation may include provider side federation and customer side federation. Community cloud the most probably will adopt federated cloud model. Intercloud. The general model for a cloud infrastructure that combines multiple heterogeneous clouds from multiple providers and typically also includes campus/enterprise infrastructure and non-cloud resources. Intercloud model may use federated cloud model or implement more specific common control and management functions to create a kind of Intercloud virtual private cloud The slide provides a visual presentation of the NIST Cloud Computing definition. It specifically reflects importance of resource pooling and virtualisation which are both the cloud properties and the essential cloud enabling technologies. Cloud resources are typically consolidated/pooled in the big datacenters running complex cloud services management and provisioning platform, also called cloud middleware. Or Cloud OS.

Cloud IaaS Properties and Characteristics This slide illustrates the five groups of the Cloud Computing properties that are generally inherited from the previous technologies: Service and Utility Computing, Service Oriented Architecture (SOA) and Web Services, Virtualization.

Scalability and Elasticity What do scalability and elasticity mean in IaaS? Clients should be able to dynamically increase or decrease the amount of infrastructure resources in need. Large amount of resources provisioning and deployment should be done in a short period of time, such as several hours or days. System behavior should remain identical in small scale or large one. How to approach scalability and elasticity in IaaS?

Transcript


For computation resources: Dynamically create or terminate virtual machines for clients on demand. Integrate hypervisors among all physical machines to collaboratively control and manage all virtual machines. For storage resources: Dynamically allocate or de-allocate virtual storage space for clients. Integrate all physical storage resources in the entire IaaS system Offer initial storage resources by thin provisioning technique. For communication resources: Dynamically connect or disconnect the linking state of virtual networks for clients on demand. Dynamically divide the network request flow to different physical routers to maintain access bandwidth.

Availability and Reliability Availability and reliability are the components of the more general property “resiliency” that means stability and ability to restore normal operation and functions. What do availability and reliability mean in IaaS? Clients should be able to access computation resources without considering the possibility of hardware failure. Data stored in IaaS cloud should be able to be retrieved when needed without considering any natural disaster damage. Communication capability and capacity should be maintained without considering any physical equipment shortage. How to approach availability and reliability in IaaS? For computation resources : Monitor each physical and virtual machine for any possible failure. Regularly backup virtual machine system state for disaster recovery. Migrate virtual machine among physical machines for potential failure prevention.

For storage resources: Maintain data pieces replication among different physical storage devices. Regularly backup virtual storage data to geographical remote locations for disaster prevention. For communication resources: Built redundant connection system to improve robustness.

Transcript


Manageability and Interoperability What do manageability and interoperability mean in IaaS? Clients should be able to fully control the virtualized infrastructure resources which allocated to them. Virtualized resources can be allocated by means of system control automation process with pre-configured policy. States of all virtualized resource should be fully under monitoring. Usage of infrastructure resources will be recorded and then billing system will convert this information to user payment. How to approach manageability and interoperability in IaaS? For computation resources: Provide basic virtual machine operations, such as creation, termination, suspension, resumption and system snapshot. Monitor and record CPU and memory loading for each virtual machine. For storage resources: Monitor and record storage space usage and read/write data access from user for each virtual storage resource. Automatic allocate/de-allocate physical storage according to space utilization. For communication resources: Monitor and record the network bandwidth consumption for each virtual link. Automatically reroute the data path when computation and storage are duplicated.

Performance and Optimization What do performance and optimization mean in IaaS? Physical resources should be highly utilized among different clients. Physical resources should form a large resource pool which provides high computing power through parallel processing. Virtual infrastructure resources will be dynamically configured to an optimized deployment among physical resources. How to approach performance and optimization in IaaS? For computation resources: Deploy virtual machine with load balancing consideration. Live migrate virtual machines among physical ones to balance the system loading.

Transcript


For storage resources: Deploy virtual storage with hot spot access consideration. Live migrate virtual storage among physical ones with different performance level. For communication resources: Consider network bandwidth loading when deploying virtual machines and storage. Dynamically migrate virtual machines or storage to balance network flow.

Accessibility and Portability What do accessibility and portability mean in IaaS? Clients should be able to control, manage and access infrastructure resources in an easy way, such as the web-browser, without additional local software or hardware installation. Provided infrastructure resources should be able to be reallocated or duplicated easily. How to approach accessibility and portability in IaaS?

For computation resources: Cloud provider integrates virtual machine management and access through web-based portal. Apply the virtual machine standard for portability. For storage resources: Cloud provider integrates virtual storage management and access through web-based portal. For communication resources: Cloud provider integrates virtual network management and access through web-based portal.

Cloud Computing restraining factors and challenges The main cloud service and operational model of hosting cloud services and resource on the cloud data center facilities rises a number of concerns and creates a number of restraining factors for wider cloud adoption. First of all, this data security and protection. The customers and users what to know where their data are stored and located and who has access to them and control over them. This is related to both company’s application data and application users’ data.

Although cloud can improve services resilience and availability, still in case of massive cloud provider outage the situation becomes out of the customer control, and this might be critical for some services and applications.

Cost is another restraining factor for using cloud based services. Cloud services cost models is seemingly simple but when it comes to running the whole enterprise IT infrastructure on clouds some cost components may increase significantly such as data upload, storage and

Transcript


access. Transition and redesign of some legacy application may also became a restraining factor.

And finally, a so-called provider lock-in factor can be taken into account when an enterprise considers which provider to select. Currently when we have quite mature cloud standardization, the majority of cloud providers comply with the main cloud standards including Application Programming Interfaces (API), VM images, storage format and others. This will allows cloud services portability and possibility to relocate them to another provider and avoid provider lock-in.

Cloud Business and Operational Models We have arrived at a summary of Cloud Business and Operational Models. Now we can define the main features – and business “effectiveness” criteria – which really define the cloud business and operational models.

First of all, the Cloud services the end user; it is a “self-service” model, making it easy to instantly sign up an provision into the cloud. It cannot be understated how revolutionary it is for the end user to simply be able to get done what he or she wants to get done without extensive process or corporate involvement.

Next, the model is through and through “pay as you go”, that is an “on demand” pay per-use model. This not only makes it low entry point and frictionless for the end user, it aligns it (assuming pricing of end user software is done correctly) to predictable margin OPEX expense; this is a crucial innovation element with respect to the business model. Another interesting and important thing which the Cloud model enforces is, that there is a clear split of responsibility between the cloud “platform” and the user, This “layer” that the user has developed with the kind of Cloud Application the users running, It is important that that be exposed as an API and that the Cloud Service provider provide as san offering at the right leveler.

Relation between IaaS, PaaS, SaaS This figure illustrates what aspects or functionalities in the general cloud services model are managed by the provider or the customer or user. The figure shows the following functional layers of the typical application:

♦ Hardware platform including CPU/servers, storage, network ♦ Virtualization layer hosting virtualization software ♦ Operating System

Transcript


♦ Cloud management software ♦ Runtime environment ♦ Application ♦ Data

Depending on the cloud service model, the layer management responsibility is split differently between the CSP and the User. The cloud services management software (or cloud middleware) depending on the cloud service model may take a place above virtualisation layer for IaaS cloud and above runtime environment for PaaS cloud. It important to mention that Data management is always remains the responsibility of the User.

Economics of Cloud Computing Now we take a closer look at the economics of Cloud Computing. First we are going to look at the way a traditional Data Center is constructed and also the operational considerations there, to understand how this rolls up into a total economic equation. Then, we will take a look at Cloud Data Centers. These differ from Traditional Data Centers in terms of several important attributes which contribute to a very different economic model. Those things include the results of Economies of Scale, the Dynamics of Supply and Demand, and also the efficiencies gained from using shared resources, which is also called “multi tenancy”.

We will take a look at these economic considerations from several viewpoints, including the Service Provider view, the customer view, and the developer / integrator view,

Cost of running a traditional Data Center Now, as modernization and overall business sophistication / competitiveness continue, the role of IT increases, IT increases business agility and makes a company more competitive. This is not only a factor for businesses but is also a factor for public sector (operation of Government) as well as in education, and virtually every organized form of work. Business climates, markets, and customers change rapidly, and rigid pre-determined IT approaches can no longer serve these dynamic environments effectively, It is now difficult – impossible –to predict and handle the demands on businesses given the digital world. Products, Entertainment, Media, all can go “viral” and create demand peaks unseen in the world until now. Traditional IT is not well suited for this dynamic environment.

Transcript


If one considers the cost of running a datacenter, it is enormous. At hundreds of thousands of dollars per rack of specialized servers and networking, including power, cooling, and networking, costs run well into the tens of millions, up to 25 million US dollars. While on the order of 40 percent is capital expenditure in terms of physical assets that leaves 60 percent for operating expenses. Of those, 2/3 is labor cost, It costs more in people than power to run a datacenter!! Now, most datacenters are very heterogeneous. They run many different applications and have a wide variety of workloads, Many of the applications in a datacenter are actually used by a few people –there are many applications, many machines, and many users, but a very low users to applications to machines ratio. Many of the servers and applications in a datacenter are old systems kept around for access to old data and infrequently used.

Cost of running a Cloud Data Center Cloud Datacenters are built using different driving factors. They are constructed such that at the physical layer they are Homogenous –all the same. At the virtualized server layer, the virtual assets can be heterogeneous –all different, to serve the portfolio of different applications. But this diversity is implemented as software that is software based virtualized resources. So they can run workloads of legacy software, and at the same time run new workloads of more “cloud native” applications – distributed and scalable web applications for example. Cloud Datacenters are built to scale differently as a result of this workload target. They are designed to handle the size of the current businesses IT workload but also the possibility of certain IT resources exploding in the business because of the rapid “network effect” companies can experience with demand (if they are lucky).

Cloud Computing instances are placed in datacenters with plenty of headroom to expand. Or a public cloud may be utilized where the headroom to expand is already massively built out and all the application developer/integrator has to do is to call upon it in the automated growth capabilities of the application This allows applications to not be constrained by the rigid IT configurations of the past; Cloud datacenters are set up to have “on tap” headroom for applications. Certainly application developers who have gone through such a “success cycle” stand by their ability to utilize smartly Public Cloud and Private Cloud together (“Hybrid Cloud”) as one of their keys to success.

The way Cloud Datacenters are constructed (as we will later analyze), the way they are operated (also) yields tremendous cost savings. Labor costs drop to 10% or less, Efficiency

Transcript


in use of the physical hardware (do more with less) increases radically due to the ability to “fill up” a server with “VM Packing”. Operations are easier using VM’s.

In the end efficiencies in Labor, Power/Cooling, and even Capital Expenses increase dramatically as can be seen.

Economics of cloud: SaaS Example Let’s examine some specific examples in SaaS and compare this with the traditional on-premise model. Assume traditional software product costs a one-time license fee of $100,000 plus an annual fee of 20 percent for maintenance and support. Costs for on-premise datacenter over five years estimated as $200,000. In addition one needs qualified people to set up all the equipment and software as well as the management and monitoring, security, and network. Today, companies do not want to spend money on expertise which is not the main subject of their business. In other words they really don’t want to hire the software and systems expertise to install and run all of this. They especially don’t want to spend the capital on all this equipment and have to replace it at each depreciation cycle.

SaaS provides a way for the company to access the application over a Browser and install/manage nothing. Essentially SaaS is a solution for non-IT companies and departments –they can use the application bit don’t have to actually run it.

The trade-off is that the company pays a subscription fee per user for the application, sharing the hardware, networking, and software which the solution provider has set up.

For example, to support 50 users, it will cost between $10 and $150 per user, per month. This includes support, general training, and data center services High-end estimate when using the CRM SaaS application for 50 users for 5 years will run about $37,500 As can be seen this results in a cost saving for the company. While each company pays less to the vendor, the barrier to entry in both expense and time for additional companies to use the application is much lower, resulting in an overall higher use-rate by companies. If the SaaS company manages acquisition of new customers efficiently and manages low rates of attritions for existing customers, overall total revenue is higher in the SaaS model than in on-premise software,

Transcript


Migrating Applications to Clouds Now let us consider what it takes to take advantage of a Cloud Computing platform in terms of moving an existing application. Some applications are structured in a manner which is more aligned to a Cloud Computing “platforming” and some are not. Some configuration work may need to be done. For example, consider applications with web front ends. With Cloud as application demand increases, one would consider increasing the number of web server “front ends” for the application in response to increased demand, Not all applications can dynamically accept reconfiguration of additional web server front ends. It really depends on the application. As another example, some applications use IP Multicast to manage a cluster of machines; in most cases Cloud Computing infrastructures do not support IP multicast in their networks. These are examples where applications may have to be restructured to be deployed in any kind of “cloud Native” configuration. If the application cannot take advantage of any of the cloud capabilities, or uin unsupported running on a hypervisor for example, rethink the motivation to move the application to the cloud! These cost factors are similar in deploying to public and private clouds. Some investigation in moving the application first to a virtualization system would be worthwhile. In fact, many organizations, worried about security or other aspects to cloud, can find a lot of utility getting started with a private cloud or groups of small virtualization cluster. This gives familiarity in managing applications as pooled virtual resources.

Evaluating applications cost in clouds Anticipate that the cloud won’t necessarily be less expensive and it won’t necessarily provide the same level of service as your data center. Your own data center may have a service level agreement with a 99.999 percent uptime record. Will your cloud provider offer that same level of service? Probably not. You have to weigh how critical that level of predictable uptime is to your internal customers. Evaluate all cost components of running applications to make a fair/comprehensive comparison: Server costs (A): With this and all other hardware components, you’re specifically interested in the total annual cost of ownership, which normally consists of the cost of hardware support plus some amortization cost for the purchase of the hardware.

Transcript


Storage costs (B): In situations where a storage area network (SAN) or network attached store (NAS) is used for an application, a proportional cost over the whole SAN or NAS needs to be determined, including management and support cost for the hardware. Network costs (C): This needs to be carefully considered because the fact that an application moves into the cloud does not necessarily mean that all the network traffic it generates disappears.

Backup and archive costs (D): The actual savings on backup costs depends on what the backup strategy will be when the application moves into the cloud.

Disaster recovery costs (E): In theory, the cloud service will have its own disaster recovery capabilities, so there may be a consequential savings on disaster recovery. However, you need to clearly understand what your cloud provider's disaster recovery capability is.

Data center infrastructure costs (F): A whole series of costs including electricity, floor space, cooling, building maintenance, and so on can’t easily be attributed to individual applications, but can usually be assigned on the basis of the floor space that the hardware running the application occupies. For that reason, try to calculate a floor space factor for every application.

Platform costs (G): Some applications only run in specific operating environments. The annual maintenance costs for the application operating environment need to be known and calculated as part of the overall costs.

Software maintenance costs (package software) (H): Normally this cost element is simple because it comes down to the software’s annual maintenance cost. However, it may be complicated if the software license is tied to processor pricing. The situation could be further complicated if the specific software license is part of a bundled deal. Software maintenance costs (in-house software) (I): Such costs exist for all in-house software, but may not be broken out at an application level. Help desk support costs (J): It's necessary to analyze all help desk calls at an application level to determine the contribution of an application (if any) to help desk activity. The support costs for some applications may be anomalous and may disappear with the movement into the cloud. Operational support personnel costs (K): There is a whole set of day-to-day operational costs associated with running any application. Some are general costs that apply to every application, including staff support for everything from storage and archiving, to patch management and networks and security.

Transcript


Infrastructure software costs (L): A whole set of infrastructure management software is in use in any installation, and it has an associated cost. TOC/App = A+B+C+D+E+F+ G+H+I+J+K+L

Evaluating Cost of Hybrid Cloud Hybrid Cloud cost evaluation requires a key understanding as to what the use of the hybrid part will be and how much work it will be to seamlessly integrate the public cloud (for that purpose) into the enterprise environment including the private cloud. First one starts out with a cost calculation on the Private cloud, and then one adds the costs of the parts which will be running in the public cloud. Are those running only when the private cloud consumes all its resources, and the public cloud is used for “peak” resource needs? Or is the Public cloud being used as a backup or replication of the private cloud computation or data.

In any case, usually the data that needs to be utilized by the public cloud part, has to be transported up to the public cloud, and there is usually a data ingestion cost both in processing time as well as in data transit. Finally is that data going to stay up there as archive for long period of time –in “computation ready” format, or as archive on a lowest-cost basis? If the enterprise requires some kind of compliance like PCI, HIPAA, etc., the use of the public clouds provide an extra layer of complication. While the public clouds have aspects of compliance, special analysis is needed to make sure one’s entire private-to-public system “in scope” meets the compliance requirements. Generally, hybrid cloud requires a greatly revised, almost new, IT policy, These are many new consideration not only from the private cloud side (because of virtualization) but because of the use of outside infrastructure –in a dynamic way. Each scenario is different, and this is a vibrant area of specific study and skill.

Economics of Cloud Computing - Government Most large governments have very large IT operations. Almost always governments operate their own datacenters. In the US the Federal Government is the largest customer for IT products.

Transcript


The numbers provided show that the spending on IT is forecast to continue to increase such that the US is planning on investing over $500B over the next 5 years in IT and IT Infrastructure.

That said the Federal government is planning a massive IT reduction, with a push to utilize more common practices and infrastructure across agencies, and seek radically more efficient ways to reduce IT spend. To that extent is is not lost upon the government that Cloud Computing provides a much more efficient use of CAPEX resource for Clouds which the government is building, to the ability to utilize “rent, not buy” scenarios, These scenarios will likely play out in roughly the order of adoption complexity for the government First governments will take advantage of existing private sector clouds (SaaS like email, IaaS like archive space). Governments will begin to tie internal architecture and small clouds, into the public clouds (eg, construct a hybrid cloud scenario). While advanced government agencies will likely to have cracked large scale cloud construction across the government gaining this skill and using larger and larger private clouds, and clouds across agencies, will be the last phase of government adoption as it is the most technically challenging. All the considerations for enterprises moving applications to clouds are present in the government scenario but “amplified”. There will be a need for governments around the world to have good skill sets in this transition.

U.S. Federal IT Market Forecast 2013-2018 The last slides have spoken to size of government IT and their eventual migration to cloud. Here is an illustration detailing the areas which the US Government will look to invest in. As can be seen, Cloud Computing and related technologies are listed as hi priority.

Business Agility and Economics of Cloud Computing (1) One of the largest benefits a user of Cloud Computing can realize is drastic enhancement to business agility. This is one of the main reasons enterprises are thinking about Cloud Computing. Business agility is the ability of a business to adopt quickly and cost-efficiently in response to changes in business environment Agility rates of the business value proposition are directly dependent on the agility rate of the underlying IT infrastructure. Applications deployed on Clouds, if properly done, are more agile than the last generation applications. They can automatically grow and shrink, they can be

Transcript


extremely reliable, and they can be changed and deployed at scale quickly and with no additional CAPEX. New versions of applications can be installed and test-driven by geographic regions.

This whole notion of continuous business transformation is a new trend in business competitiveness. Cloud Computing is a strategic weapon capable to enable full business transformation (including operation)

Business Agility and Economics of Cloud Computing (2) Clouds bring IT agility in economically efficient way. IT can use On-demand services to pay as they go, or use predictable OPEX pre-established costs of using compute and storage resources. These approaches enable agility as it allows the enterprise to make IT changes – new applications, changed applications, changed data (like pricing) –and that’s called agility. But IT shops don’t operate in a way conducive to this model. Application owners change from becoming customers of IT to becoming service providers of their application. They are provided with a homogeneous automate IT platform (Cloud) and their app/service goes on top of that. This is not a current approach in an environment of heterogeneous silos of technology. There is a transformation required in IT and with the applications owners to get to cloud.

Cloud and Total IT Spending In this study on “Business Agility and the True Economics of Cloud Computing”, their research shows savings from various styles of usage of Cloud Computing. The analysis shows that more sophisticated use of Cloud Computing (Hybrid) can yield the largest efficiencies.

Factors of Cloud Economics This chart shows the reasons why Cloud Computing can yield savings and efficiencies. It shows that in comparison to existing Datacenter technologies, Cloud has three main architectural advantages: 1. Supply-side savings yield large economy of scale. 2. Demand-side aggregation: Aggregating demand for computing smooth's overall variability, allowing server utilization rates to increase. 3. Multi-tenancy efficiency

Transcript


The combination of these efficiencies compounds into the Cloud platform, giving it characteristics which are superior in efficiency well past that achieved by previous computing architectures. The next slides go into more detail on each of these individual characteristics.

Supply-side savings: Economy of scale Economy of scale is all about costs of goods going down (on a per unit basis) as scale increases, in terms of what are COSTS (supply side) to offer the service. Businesses which exhibit this characteristic quickly find that larger suppliers, the ones best positioned to run at high scale, quickly dominate the market. This is precisely why we see most Cloud Computing players emerging as large providers. Power is the second largest expense in running a Cloud next to operational staff. Power effectiveness is becoming very high with today’s computing environments. Therefore, one sees large datacenters containing Cloud Computing located near inexpensive energy sources. As mentioned, operational staff is a large expense, Because Cloud automates the spinning up, spinning down, and many maintenance issues surrounding computing, the need for manual processes is decreased. Therefore the same size staff can run a much larger installation. This leaves human budget for creating new services, delivering customer service, and other higher value activities. While one of the top concerns in Cloud Computing is Security, in actuality, Cloud installations are naturally safer locations than enterprise or special purpose datacenters. A larger operation overall has a larger budget for Security for example. The Cloud provides a larger “attack surface” and so more attention to security is required at a Cloud installation. Overall more attention and more efficient policies and processes are in place at Cloud locations than other datacenters. Finally, huge equipment and software discounts are had by Cloud operators who buy thousands of servers and switches or software modules at a time. They enjoy the largest of “volume discounts”.

Demand-side Economy of Scale This slide speaks to Demand side Economy of scale.

Transcript


Demand side speaks to all of the dynamics and behaviors of how the system is used which add to the economy of scale. Consider each of the following examples, they are all following the themes of “statistical multiplexing” of the users. Randomness speaks to the spreading load out over time from a large user community. The larger the random set of users the more the use pattern becomes less “spiky” and more demand-consistent. The same happens in Time of Day statistical multiplexing; a large number of users make the demand seem consistent through time. The other examples -Industry-specific variability -Multi-resource variability -Uncertain growth patterns: are all driven by industry dynamics: all show how diverse factors on the demand side lead to an overall evening out of the “total” demand on the system. This is a very important concept, overall the Cloud –if used in enough diverse ways by a large enough community – will in the aggregate appear to be a very consistently used system. This is called Demand side Economy of scale.

Multi-tenancy Economy of Scale In another dimension of Economy of Scale comes from the use of shared (multi-tenant) infrastructure. When organizations are using their own servers, even their own virtual servers, each of them need to worry about updates, incident management, and security to name a few examples. In a multi-tenant application like most cloud based SaaS applications, updates to system software, managing incidents, and security functions are handled by one team for all tenant organizations, increasing efficiency and leading to what we call Multi-tenant Economy of Scale.

Overall Impact The overall impact of these economies of scale, as can be seen by the illustration, shows that the combination of supply-side economies of scale in server capacity (amortizing costs across more servers) and demand-side aggregation of workloads (reducing variability), and the multi-tenant application model (amortizing costs across multiple customers) leads to powerful economies of scale. The model indicates that a 100,000-server datacenter has an 80% lower total cost of ownership (TCO) compared to a 1,000-server datacenter.

Transcript


IT Spending Breakdown Why are considerations like Economy of Scale important to consider anyhow? This illustration shows that a large portion of IT spending is actually on infrastructure. More than half of IT spending goes to “plumbing”. If Cloud Computing can reduce the spending on infrastructure to almost nothing, it can essentially cut IT costs for a given amount of application software –in half.

Or put another way, if one can get infrastructure for almost nothing, for the same overall budget, 4 or 5 times as much attention can be paid to applications development.

Cloud Economics/Benefits are Good for existing Applications, are Great for new Applications How much actual economic advantage can be had from Cloud Computing, depends largely on how much one can leverage those benefits in the application. Moving existing applications to cloud may have certain limitations which prevent them taking full advantage of what Cloud can offer. As the slide details, certain architectural issues in the application may prevent one from realizing the auto scaling, or self-healing, that Cloud can offer.

In this case, one might consider using SaaS applications fulfilling the need rather than porting legacy applications to the Cloud.

New or custom applications, if properly architected and deployed, can take full advantage of Cloud, if they are structured correctly and use a Cloud specific deployment model.

If the Cloud offers a PaaS model, where the PaaS is “native” to that cloud, this is likely to be the absolute best way to move an application. A Cloud Service Provider integrated PaaS can provide deep integration with the benefits of Cloud. Now that standardization of PaaS environments are beginning to come along, there will be more and use of this model.

Cloud as IT Innovation facilitator: Possibilities Cloud Computing can be more than just a money saver for IT departments. Many people believe it is becoming a catalyst for innovation.

Transcript


As society gets more and more sophisticated, all businesses rely on IT to rapidly accomplish complex tasks that were previously prohibited by cost or time constraints. Cloud Computing effectively amplifies this capability. It allows for a new class of experimentation and entrepreneurship, were barriers to entry are low and where rapid expansion is possible. Cloud Computing both lowers the costs of starting an operation and lowers the cost of failure or exit. All of this adds up to Cloud as IT Innovation facilitator.

Cloud as IT Innovation facilitator: Obstacles While enthusiasm around Cloud Computing as a catalyst for innovation is high, there are plenty of obstacles for Cloud Computing. Because Cloud is not just a technology shift, eg, it includes a change in business model (pay as you go), and it often involves a change of control (data and computing now happens off premise at a Cloud Service Provider one really knows very little about). As a result security, privacy, maturity, and compliance are the top concerns, followed by concerns about legacy (backward) compatibility. As this slide details, there are additional concerns.

There is a concern over being ever to “go back” from Cloud, once your data and your application have been moved, is one stuck with Cloud, unable to return to the “old” way of doing things?

The complex security mechanisms tried and true in traditional computing don’t automatically work for Cloud systems and applications, where virtualization, “as a service”, loss of control, and use of shared resources are all standard in the Cloud.

Additionally, there are concerns over platform maturity, Availability and Performance, and Compliance and Data Sovereignty.

What is the way to go? So, given all the concerns, is Cloud Computing the way of the future? It is clear that the Traditional Data Center will definitely undergo transformation to private cloud platform keeping some legacy facility and applications.

This Private cloud will provide benefits and economy in simplifying IT infrastructure management and maintenance

Transcript


Many people will find the Public Cloud attractive for all kinds of reasons, ranging from low cost for the smaller businesses, to any size businesses which want to be able to scale up and serve millions of people with only a small group of developers. Of course, there is a Cloud topology which is somewhat the best of both worlds of Private and Public. Hybrid cloud provide benefits of both private and public clouds and are optimal for medium sized and large companies, who also owning own data centers. Hybrid clouds bring maximum benefits for modern agile companies.

Summary and Take away Cloud Computing has many benefits as a new technology and as IT infrastructure design and management transformation factor At the same time Cloud Computing and a number of restraining factors, main of which is security of data and services or infrastructure in clouds Cloud Total Ownership Cost (TOC) structure is complex and not as simple as presented by the Cloud Service Providers. Besides costs of running services and applications in clouds, it includes also include transition costs, possible applications re-design, and in-premises IT management For hybrid clouds TOC includes both cost of private cloud and services outsourcing to public cloud Cloud Economy of scale comprise of 3 factors: demand/customer side, supply/provider side, and multi-tenancy Moving enterprise IT infrastructure bring an agility both for IT infrastructure and to company itself Cloud as IT innovation facilitator can free significant resource in company which can be used for the company’s main business.

cloud benefits and challenges/problems, business and operational...

Documents