Profile Facebook Twitter My Space Friendster Friendfeed You Tube
Kompas Tempo Detiknews
Google Yahoo MSN
Blue Sky Simple News Simple News R.1 Simple News R.2 Simple News R.3 Simple News R.4

Sunday, March 13, 2011

Cloud CONTROL - Managing Cloud Services Demands A Different View From IT

Like a helium balloon filled by an overconfident employee, the cloud is often pumped with hot air by vendors and other proponents looking to overstate its value. But while the benefits of the cloud might be overblown at times, there’s no denying its technologies are having a profound impact on businesses that choose to use them.


That impact typically contains a nice selection of advantages, but it also places challenges squarely on the shoulders of IT departments.

The cloud can make the jobs of IT employees easier, but it requires a unique approach to in-house IT operations to ensure its benefits are delivered as intended. IT departments that make either a full or partial transition to the cloud ultimately must expect to alter their management style to accommodate cloud-style requirements that tend to contrast significantly from in-house requirements.

Changing Roles
Ask anyone in the trenches to gauge the cloud’s impact on IT departments and the answer will likely be: “It depends.” More to the point, the impact depends on the type of service a business uses. For example, says John Barnes, chief technology officer of Model Metrics (www.modelmetrics.com), the impact delivered by a SaaS (software as a service) solution can be fairly limited with the exception of user and identity management or possible integration with a single sign-on product or a back-end legacy system. But other cloud flavors can have far-reaching effects on IT organizations.


“If the cloud services fall into the PaaS (platform as a service) category, the company’s IT department will experience a big weight lifted off its shoulders,” Barnes explains. “The solution will take away the heavy lifting that an IT department needs to worry about when deploying a new solution. It will allow them to focus on the development and real business value and let the PaaS provider worry about the infrastructure, the patch management, the clustering, and disaster recovery—all things that distract IT departments from more strategic projects.”


The benefits can pour in even heavier when a business travels the IaaS (infrastructure as a service) route. David Powers, senior analyst of business development at Cycle Computing (www.cyclecomputing.com), says that IT departments using IaaS will experience a positive disruption of sorts in terms of the time to provision a machine or cluster of machines, transparency of resource usage, transparency of cost at a very fine granularity level, elasticity in terms of the number of machines deployed, the ability of Web service calls to execute highly complex infrastructure configurations and deployments, and the ability of end users to enjoy self-service capabilities.

“All of this is possible without ever having to purchase hardware assets in the data center,” Powers adds. “The cloud services are all pay-as-you-go, such that you only pay for what you use, and when you’re done using it, you stop paying for it. That is very disruptive to how enterprises have historically deployed hardware assets in their internal environments. Cloud computing does not necessarily mean an end to traditional system, network, and storage administrators. Rather, what it does mean is a change in skill sets that folks in these roles have traditionally had.”

Traditional in-house architectures require IT personnel to have low-level, nuts-and-bolts expertise with systems, networks, and storage, but cloud deployments can remove at least part of those requirements. In turn, Powers says, this lets IT departments move their focus higher up in the stack to the application and analytics level, where they are of significant value to the business as a whole.

Watch The Workflow
The service concept ingrained in the cloud inevitably becomes part of any business that opts to use a cloud product. But there’s a side effect to this new operating model, says Mark Thiele, vice president of data center strategy at ServiceMesh (www.servicemesh.com). Through the cloud, IT becomes “just-in-time” IT, where formerly complex, high-risk, time-consuming projects become just-in-time services that are available when and where customers need them.

“Historically speaking, applications have been adopted and maintained based on a very long cycle of review, ROI investigation, architectural fit, resource availability, and other factors,” Thiele says. “The timeline for the aforementioned cycle is often measured in at least quarters and sometimes years. A side effect of the current application ownership strategy is that you create a large underlying infrastructure pool that has stranded capacity in space, hardware, middleware, skill sets, and [operating expenditures],” Thiele says.

By integrating cloud services, IT budgets can move from a split of 75% maintenance and 25% innovation to something along the lines of 50/50 or even 40/60, favoring innovation. Thiele notes this will free up both capacity and funds to allow IT departments to focus on solutions that can help add value to the bottom line. Thiele also says that IT departments moving to a service-oriented architecture might find weaknesses in their workflow processes and capacity planning strategies.

“Workflow response in today’s IT organizations is measured against historical fulfillment cycles of days, weeks, and months,” Thiele says. “In an everything-as-a-service model, the workflow will have to be modified to take into account the faster fulfillment cycles. IT will also need to be adapted to new providers, both internal and external. Capacity management will become the skill of the decade for companies with large cloud environments. Without effective capacity planning, you run the risk of contracting for services, scale, and availability requirements that either don’t meet business requirements or exceed them.”

Read The Fine Print
After years of interaction with in-house infrastructure and technologies, IT staff naturally develops techniques and management styles designed to take advantage of these pieces. The cloud can throw a big, shiny wrench into the machine, but it’s possible to leave that wrench where it falls as long as staff can adapt to its presence.

“When you have spent years working handson managing systems and now you have to open a ticket with someone else, it can be a bit stressful initially until you get a routine down,” says Mark Gilmore, president of Wired Integrations (www.wiredint.com). “Don’t expect things to happen as quickly as they used to, and be prepared for the fact that you are no longer in control of the service being outsourced.”

A critical piece of advice that Gilmore lends to his clients is to read the fine print of their service contracts. Many people, he says, are shocked when they encounter an issue and cannot manage that issue like they used to due to restrictions in the service contract. However, if businesses learn up front about how cloud service providers operate, they will be better prepared to interface with them when issues arise. Also, it’s important to never assume that everything will always be status quo with providers—because that’s not the case.

“I had a services firm client that used a file storage service and was under the assumption that they provided 24/7 service, until the day came on a weekend where the service went out,” Gilmore relates. “After doing some research, I discovered that this particular vendor only operates on a normal business day schedule and provides no after-hours support. The client was stuck and lost several days of productivity due to the service outage,” Gilmore relates.

All Together Now
Businesses that transition multiples services to the cloud might find themselves struggling to manage these disparate parts, but Thiele recommends adopting a management and orchestration framework that lets IT treat all of the services as if they are one. This “single pane of glass” approach can help to lower the cost of ownership, reduce the risk of vendor lock-in, and ensure greater governance, policy enforcement, and applied automation. Another way to view the services is to consider them as LEGO blocks.

“At the IaaS layer, you might think of the cloud services as LEGO blocks, and someone inside the internal IT department will need to be able to architect and design the infrastructure LEGO blocks in such a way that they provide value,” Powers says. “A bunch of disparate LEGO blocks don’t really add much value. But when assembled, those LEGO blocks can provide very powerful computing environments. As a result, the traditional role of system administrator will require the skills to assemble disparate LEGO blocks which are already provided to you and stack them together to make robust, flexible, scalable, and secure computing environments that optimize price, utilization, and performance.” ▲

Related Post:

Widget by [ Iptek-4u ]

0 comments:

Post a Comment

 
Copyright © 2010 - All right reserved