Archive for the ‘Data Center Outsource Costs’ Category
About two years ago, the federal government set in motion an ambitious and broad initiative to dramatically reduce IT operations. The goal: slow the sprawl of the government data center footprint.
Well, current federal CIO Steven VanRoekel says the plan is exceeding original plans and targets. In this blog on the White House’s Office ofManagement and Budget website, VanRoekel says by the end of 2012, the federal government will have closed 740 data centers. By the end of 2015, it will consolidate at least 1200 data centers, or at least 40% of identified data centers – a goal that VanRoekel writes “requires us to continue aggressively rooting out duplication and waste in our expanded baseline of 3,133 data centers.”
The data centers comprise all sizes, ranging from some as big as a football field to others as small as a closet. They represent, VanRoekel reports, billions in wasted capital.
The consolidation and closures are expected to save taxpayers billions of dollars by cutting spending on underutilized hardware, software and operations; improving cyber security; shrinking energy and real estate footprints; and taking advantage of innovative technologies such as cloud computing.
The government recently expanded its initial plan, and in the fall VanRoekel and team announced plans to include in the consolidation initiative data centers of any size, not just those that are 500 square feet and above. Moreover, the government says it is paying close attention to areas within the remaining data centers where greater efficiencies can be realized.
More of the ITWorld article from Beth Bacheldor
Consider how Cloud Computing’s unique characteristics will change how you do architecture
Today is a wonderful time for anyone interested in Cloud Computing to be working with the US government. On the one hand, the government considers Cloud to be strategically important, and they already have a track record as an early adopter of Cloud Computing on a grand scale. On the other hand, the government is also in the unique position of being able to drive standards for the approach—and in fact, they are even responsible for establishing the most widely adopted definition of Cloud Computing.
The federal agency who has taken this leadership position is the National Institute for Standards and Technology (NIST), an agency of the US Department of Commerce. NIST’s formal definition of Cloud Computing is already well known—“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.” Concise as that definition is, it only marks the beginning of the work NIST is doing to formalize and standardize the full breadth of Cloud Computing approaches, both within the government as well as for the world at large.
More of the Cloud Computing Journal article from Jason Bloomberg
99.995% uptime and affordable colocation are not mutually exclusive. Many companies that visit our data center are surprised to learn that they can have high reliability without the huge capital costs of building a tier IV data center.
Most companies who need new high uptime data center space compare the costs of building their own primary data center in a company building versus using wholesale colocation facilities, aka outsource data centers.
So what does high uptime mean? Uptime is the measured value in minutes of a company’s computer systems reliability. 99.995% uptime means 28 minutes of downtime per year or less. Companies who value uptime know that downtime causes lost sales, lost profits, and lost clients. These companies haveoften learned about the costs of data center downtime the hard way. Some unlikely circumstance caused an outage that was painful enough for leadership to reevaluate the importance of the server room to the success of the company.
But the cost of uptime is high. A small in-house data center with 99.995% uptime can easily reach $1 million in capital costs, and tens of thousands in staffing, yearly maintenance, SAS 70 and SSAE 16 data center certifications.
What does affordable mean? Here are three characteristics:
Simple data center pricing model – Can you understand how the pricing works? Are there multiple add-on charges and mysterious extra monthly fees?
Predictable – Predictable pricing models make it easy to forecast growth and change. How complex is a three year analysis of your costs? Are there multiple variable costs?
Incremental- Incremental means pay as you use it. Can you grow the number of racks and pay accordingly? Do you pay for electricity as you use it, or based on the circuit size?
The good news: you can have your cake and eat it too. You can meet tier IV data center uptime requirements and still keep data center outsource costs low. The bad news: there are only a few Midwest colocation facilities that offer high data center uptime at affordable pricing. Do your homework and you’ll find flexible affordable colocation with high data center uptime.
Enterprises looking to revamp their strategies around data center optimization and utilization face the initial decision of whether to build from scratch or to retool their existing systems.
Creating a new data center from scratch allows an organization to plan for the greatest amount of customization. However, the significant investment in time, construction, hardware and support is prohibitive for many enterprises. Reconfiguring an existing data center is efficient and cost-effective, but it cannot always fully address an enterprise’s requirements.
More of the Baseline article from Adam Wallace
Is a simpler data center pricing model better? Is the data center pricing model itself a decision factor when companies are reviewing colocation? We believe the answer is yes. Let’s look at the history of outsource data centers, also known as colocation, for some perspective.
In the early days of colocation (the mid-1990s) many colocation providers grew out of the telecommunications space. A Director of Operations at a local telecom branch office probably looked at some empty space in his building. He then called the Sales Manager and asked if he could find a client who might be interested in renting the space. The Sales Manager found a client, so they had to come up with a pricing model. As many of us already know, telecom providers have some of the most complex and convoluted pricing models of any vendor. Many of the complexities of telecom pricing models came to the colocation world. Colocation seems to have been born with a complex pricing model.
How many line items are on the typical monthly colocation invoice? I’ve had clients who use other data centers tell me that they have 10 line items to rent a single cabinet and a little bandwidth.
But an outsource data center is really not that complex. All colocation facilities provide real estate, power, cooling, and access to bandwidth. Midwest colocation provides hardened data center facilities and F5 tornado resistant data centers. West coast colocation is often earthquake resistant. The offerings seem simple enough. Why should the colocation pricing models be complex?
Here are some features to look for in a pricing model:
How is real estate delivered? Many outsource computer room providers dictate the amount of floor space each rack is allocated. Some allow you to purchase extra space for a less dense footprint or for growth over time. Does the data center give you the flexibility you need to grow and change?
Is the power pricing based on actual draw? It is common for outsourced data center providers to bill based on a circuit size rather than the actual power used by a client. This circuit size billing method is inherently inequitable, because power needs shift over time, and circuit utilization is never more than 50% in a highly reliable data center. Look for pay-as-you-use-it power pricing.
How is the power for cooling calculated? Data center equipment (servers, network gear and storage) require about 1 kW of cooling for each 1 kW of power to operate the equipment. Does the pricing model charge you for the cooling power in a sensible manner?
How is the capital overhead of generators, UPS systems, HVAC charged? Every rack in every data center uses a portion of the power and cooling infrastructures, along with the staffing and the data center compliance overhead. Are you being charged fairly for your share of these complex and expensive infrastructures?
Does the colocation provider charge monthly cross-connect fees? Many data centers offer access to multiple carriers. But most charge you a monthly fee for the privilege of connecting to these carriers. A few data centers charge no cross connect fees. This can be a huge savings over time, especially when companies employ multiple carriers in a complex wide area network.
Use these features to compare data center outsource costs. A simple data center pricing model lets you understand what you’re spending, better forecast changes, and control the overall cost of operating your data center.
Great set of questions from last summer on using outsource data centers and the decision makeing process to do so.
I’m interested in hearing how organizations have come to the conclusion that this was their best alternative to expanding data center capacity?
We have seen some organizations dive into Co-Lo facilities because the industry is growing so they assume it is the right thing to do or they think it will be more cost effective before anyone has really analyzed the cost implications. I think it is critical for an organization to analyze all of the suitable approaches to their need of expanding data center capacity such as; upgrading existing facilities, Co-Location, building new or building a scalable Data Center Shelter that can be leased to provide tax benefits while the lease payments may be close to what you’d pay for suitable Co-Lo space.
First of all, what is driving your interest in Co-Location is it the costs of running your own facility? Or one of these issues:
More of the Data Center Design blog post
1. Choose a Top Quality Internet Network – A worldwide Tier 1 International fully redundant OC192 backbone with additional 10 GigE network connections to hundreds of other Internet networks is the best service a business can acquire. Ask the colocation provider(s) that you are considering about their Internet network connection size and network details.
2. Choose a State-of –the-Art Class A Colocation Facility – A facility that has highly scalable and super fast connections to the top Internet backbones, redundant UPS and Prime Source type of generator backed electrical power, redundant A/C systems, 24/7 on-site technical support and physical security.
3. Does the Colocation Provider Include Remote Hands for Free or for a Fee? – Ideally, you want to find a provider that does not charge for remote hands service because it can be very costly. There is no need to pay a fee when there are state-of-the-art providers that offer it for free. These colocation providers who include remote hands service for free often have faster, more responsive and experienced technical service personnel who will be there around the clock, when you need them most.
More of the ColocationProvider.org post
Data center construction costs average $295 per square foot ($150 to $200 per SF shell, $12M to $18M per MW thereafter depending on the required design resiliency) – only hospitals compare in their costs to construct. With so much cost in every square foot of a data center, it is essential to extract any potential value out of the real estate. One of the easiest ways to create this value is to a pursue cost segregation analysis.
Cost Segregation Fundamentals
The Internal Revenue Service (IRS) requires commercial buildings to depreciate using a 39-year straight-line method. For example, a $39M data center can be written off at a rate of $1M per year over 39 years.
However, through a cost segregation analysis, a significant portion of the building costs can be reclassified from “real property” to “personal property” thus changing its tax depreciation life from 39 years to 5 years.
More of the CBRE post from Eli Varol
Insufficient management tools mean staff spend too much time managing servers
The IDC survey of 300 large European businesses found that one in four organisations were managing their servers and storage manually, leading to much higher costs compared with organisations that used some tools.
Only 14% of organisations had a fully integrated management framework. The research found that only 30% of companies saw datacentre operational cost as a priority, with 25% concerned specifically about software licence costs.
Nathaniel Martinez, programme director in IDC’s Systems and Infrastructure group, said, “Datacentre managers are much more concerned with finding suppliers that can address the security and availability problems they are experiencing than with ensuring that their datacentre meets the requirements of their business.
More of the Computer Weekly article from Cliff Saran
Google’s top energy executive has offered some simple steps for making data centers more energy-efficient, including raising the thermostat to 80 degrees Fahrenheit — or 27 degrees Celsius — to cut down on cooling costs.
Data center staff at some companies walk around in jackets because the buildings are kept so cold, said Bill Weihl, Google’s “green energy czar,” at the GreenNet conference in San Francisco on Thursday. “In our facilities, the data center guys are often wearing shorts and t-shirts,” he said.
The tips he offered have been batted around at data center conferences for a few years, but it’s likely that many companies still aren’t making use of them — especially to the degree Google does at its own tightly-run facilities.
By taking fairly basic steps, most data centers could lower their PUE to 1.5, Weihl said, compared to an industry average of 2.0 or more. PUE, or Power Usage Effectiveness, measures the total energy consumed by a data center against how much actually reaches the IT equipment.
More of the Network World article from James Niccolai
