Data Center Connectivity Gets a Boost Heading into 2013 – Sidera and Lightower to Merge

- Ilissa Miller, CEO of iMiller Public Relations, says:

On December 27, 2012 Berkshire Partners, a Boston-based investment and management firm announced that it will acquire and merge Sidera Networks and Lightower Fiber Networks – two companies that operate and manage substantial networks throughout the Northeast, Mid-Atlantic and Midwest.

The definitive agreement was signed by Berkshire Partners with Pamilco Capital, a significant investor in Lightower and ABRY Partners, a significant investor in Sidera.  Since the two companies are being acquired, in essence simultaneously the deal is valued at over $2 billion.  

What Will the Combined Network Look Like?

Lightower’s current network reaches its most southern point in Princeton, NJ.  Combine their network with Sidera’s, customers will gain access to additional   diverse and ultra-low latency network connectivity that extends to the Washington, DC metro area including Ashburn, VA, as data centers in the region know – it’s the US-based media and content data hub.  Sidera’s network will also offer Lightower customers low latency connectivity to the Chicago/Aurora, IL financial exchange market as well as connections to the financial markets in Toronto, Canada and London, UK.

Sidera’s over 13,500 rout miles, combined with Lightower’s over 6,600 fiber route miles, will offer customers more than 20,000 route miles of network connectivity and provides access to more than 6,000 on-net locations, including commercial buildings, data centers, financial exchanges, content hubs and other critical interconnection facilities.

Data Center Connectivity

In 2012, Sidera Networks rolled out and executed against an aggressive strategy to connect as many data center and colocation facilities as possible.  As of the date of this announcement, Sidera interconnects over 125 data centers across its vast footprint.  Lightower’s footprint reaches approximately 64 data centers – with some over lap, it appears that Sidera’s network will provide the greatest reach to the ultimate combined entity. 

What does this mean for data center providers?  Not much more than a consolidation of suppliers.  Expect some redundancy and network optimization to take place as the companies evaluate the overlapping footprints of their networks.   The merger is expected to create an entirely new entity – so the administration aspect of your customer relationship may change. 

For more information about how this could affect your relationship with Sidera and/or Lightower, visit Sidera Networks’ FAQ page here: http://www.sidera.net/lightower-sidera-merger/frequently-asked-questions/.

API Aggregators: The Alpha Males of Data Integration

- Cyrus Ghalambor, founder of Adigami, (www.adigami.com), says:

It’s 8 PM at work and your client has asked for the latest data in the campaign you’re running for an online home goods site. She has also asked you to include one new data source you haven’t worked with before. Do you try to figure out at the last minute what API they are using, then work frantically to try and pull in the data so it is congruent with the rest of the report?

Enterprises are coming to expect more from marketing and business analysts charged with keeping their eye on the ROI. In-house digital marketing departments and external agencies also are tasked now with far more intensive analysis.

In meeting demands for more data, enterprises face the issue of APIs, the gateway to analytics data sources. APIs can stop the train since there is no commonality among them, leaving IT people to face the problem of obtaining data from sources with disparate query language.

Since lean staffs are the norm in this economy, the task of integrating this data is not a welcome one. It takes considerable time to develop connectors so your dashboard can pull in the data. At the same time, companies are no longer satisfied with hand-made Excel reports, and they can no longer spare the time for the manual download of results. They also cannot afford the errors that inevitably happen when you assemble and process so much data by hand.

The answer is the automated collection and integration of data via the cloud, but how to solve the issue of API integration? We pose the concept of an API aggregation service – the alpha male of data integration. This would implement a ‘unified API’ that provides a single interface across many APIs, eliminating the need to build separate APIs, saving your IT and engineering personnel countless hours of development and integration time.

In the digital marketing arena, using one example, the unified API is an attractive solution for cross-channel reporting. Cross-channel marketing is now the norm; analytics must follow with data from search (PPC), display, social, email, and other sources. You can see the challenges this creates when integrating data – even before analytics professionals can do their job of viewing and revising strategy.

If you’re building an app (for web or mobile) that needs to bring in data from a variety of data sources such as Google Analytics and AdWords, Microsoft Bing/adCenter, Facebook (social and ads), Twitter, DoubleClick (and others) then you’ll need to learn the API’s for each and every one of these sources and continue to stay on top of the changes they regularly make.

A unified API solves this dilemma by providing a dozen or so standard calls that are applicable across many platforms. Another benefit: API aggregators keep track of any API changes and improvements that the data providers make, ensuring that enterprises need not worry about any bumps along the data integration road

As an emerging technology, we believe a unified API is a viable and affordable solution for marketing and business analysts – and their IT support – who are struggling to find an easier way to bring in data from a growing number of sources. There is some movement in the industry for API standards but history shows that standards battles tend to drag on. A unified API – via the alpha males of API aggregation -- is a workable solution today.

Data Center Efficiency: It’s Time We Noticed The Elephant In The Room

- Liam Newcombe, CTO of Romonet (www.romonet.com), says:

At the moment, regulators across the world are struggling to find ways to measure how energy efficient a data center is by measuring how the ‘useful work’ it delivers compares with the energy consumed. The idea being that data centers can then be given a ‘score’ which says how energy efficient they are – or are not – and the regulators can then influence the market.

Whilst I agree that data centers need to become more energy efficient, I think that in the quest to find one, all-encompassing energy efficiency metric, we’re being side tracked from the main issue and ignoring the big opportunity. For me, the elephant in the room is that data center environments have traditionally been designed, developed and managed as if their workload is static – and that is what we should be concentrating on first before we start imposing a metric.

To explain this, let’s draw a comparison between how a desktop PC or laptop uses power and how a data center uses energy. In recent years, chip vendors have worked hard to make laptops and PCs smarter in the way they use power – laptops and tablets in particular need a long battery life otherwise they won’t sell. This means that if you’re doing something quite complex on your PC – like editing a video or watching a movie – the PC draws more power than if you’re just browsing the web or working on a word document – and in standby mode, a PC hardly uses any power at all. Unfortunately, data centers don’t work like this.

Let’s say we have a corporate data center which hosts a customer contact platform with a web presence and a few thousand call center operators. On a Monday morning, it might be handling 50,000 calls and 100,000 web users every hour. It might be using a lot of power – but it’s doing a lot of work. However, if we return and measure energy consumption at 4am on a Sunday when there are no phone customers and only a few web users – it will be drawing almost the same amount of power.

To use an analogy, if a data center was a car we would not be able to turn the engine off and it would use almost the same amount of fuel idling on your driveway as if it were carrying five people, speeding along the motorway at 70 MPH. To continue the analogy, if you were concerned about your car’s fuel consumption – which you really should be – you’d probably ask why you couldn’t turn the engine off and why your car needed all that fuel when it was stationary. You wouldn’t ask the garage to tweak the engine to make it run more efficiently at 70 MPH – but this is the equivalent approach our regulators are taking in searching for a metric to measure how efficiently data centers run at maximum load.

For me, it is less important how efficiently a data center operates when working hard on a Monday morning. I am far more concerned that we understand how inefficient it is when it’s doing nothing – and that we then minimize this inefficiency. However, to date, the data center industry has a poor record of coming together to address this challenge.

Whilst some companies have recognized this issue and started making more energy efficient servers, much of the equipment in a data center still draws close to full power all of the time. There’s no one single piece of equipment that’s going to solve this issue and no one vendor or department that is solely responsible – there’s no magic bullet. To address this issue we need a paradigm shift in the way that we as an industry think about the design and deployment of data centers. From the mechanical engineering to writing our software, we need to ask: “how much power, data center capacity and cost will this waste to do nothing?” And as we ensure every device is tuned to manage its energy consumption at a lower load, we’re also going to see data centers becoming cheaper to run and more efficient at full load.

But to spur the industry into action, we need to create an informed market. We need to make the issue of waste energy and cost clear to the business. Every finance chief needs to know how much energy each data center service is using when it’s working at full capacity as opposed to when it’s idling for them to be able to make a decision which service is worth the cost – and which isn’t. Only then can they make an informed choice about which data center services they are happy to spend the money on, which ones they might want to spend less on and which can be turned off altogether. And it’s only when this happens that the market will start to change.

Data Center Metrics: PUE/DCiE Confusion

- Larry Vertal, the Executive Director of The Green Grid (www.thegreengrid.org), says:

The release by The Green Grid of PUE & DCiE, which have become common measures used globally, certainly showed the pent up demand for metrics that were not self-serving marketing tools!

While The Green Grid defined the metrics carefully, provided the guidance on usage in straightforward whitepapers, released a tool to assure consistent reporting there are still claims of confusion.

While these initial metrics by TGG were intended to enable continuous improvement of a given data center, some of the stir and alleged confusion lies in desire to use the metrics to compare one data center to another. For example comparing a web page hosting data center with one focused on financial transactions that have different design points and requirements. In our competitive industry the desire to use any metrics for marketing initiatives is understandable and in the heat of competitiveness can lead to misuse and drive confusion.

I would suggest that real rather than alleged confusion in many cases comes from the fact that many folks have heard about PUE but have not been able to take the time to read the short and straightforward whitepapers of TGG defining and guidance on implementation. It is like the old children’s “telephone game” where the actual meaning and message get muddled at each step of discussion by secondary and tertiary sources. It is understandable especially with small and medium sized IT departments who often time depend on third parties for information and feel that they do not have time to go to the source but in reality the source documents in this case are quick reads!

Without taking the first steps to measurement you are unable to measure the affect you have when you implement changes or indeed do planning which is more than “dart throwing". However, it should not be so. The Green Grid defined PUE usage so it could be implemented at different levels of precision and measuring at the least precise level can give small and midsize data centers valuable information to control their destiny. At Level-1 the operator merely measures facility power at one point, the overall power going into the facility, measures the IT power used via UPS output and does the measurements monthly or weekly.

What’s a key goal? Not just on one hand how much energy is being used or on the other hand how much real work of value is being done but the combination of the two: Productivity. And measured in a manner that is do-able, of value and cannot be “gamed.” This is not easy.

The Green Grid Data Center Energy Productivity (DCeP) metric is very complete, very accurate and one of the US National Laboratories, PNNL, has worked with The Green Grid and published results that prove its validity. However because of its accuracy it can require a level of instrumentation that is a barrier to most data centers in implementation today. As the level of data center instrumentation evolves and improves generally The Green Gird is approaching this issue in two key ways. We have begun work in testing self-reporting data center applications which report out real work which when combined with power and energy measurements allow real insight.

Additionally we are involved in evaluation with the industry and end users of Proxies that may be used for energy-based productivity estimates. These include such proxies as those based on DCeP sample workloads, bits per KWH and CPU-utilization such as SPECint_rate, SPECpower and CUPS.

While DCeP is the general solution and self-instrumented applications will take time for development and deployment by vendors and users, the interim approach with Proxies is underway. Use of the proxy approach has the potential to help small and midsize data centers in particular by having low barrier to implementation.

Data Center Metrics: PUE and Energy Consumption

- W. Pitt Turner, Executive Director of the Uptime Institute (www.uptimeinstitute.org), says:

The PUE (power usage effectiveness), which is the ratio of total data center energy consumption divided by IT energy consumption, demonstrates the portion of energy at a data center required to support power, cooling, and other non-IT loads at the site. This is a measure of the ‘overhead’ required to support IT.

PUE is useful when the total energy for the data center can be easily identified. It becomes more complicated when a data center is in an office building or when the meters necessary to identify the energy consumption are not available.

The number of racks receiving cooling air outside of the ASHRAE (www.ashrae.org) temperature range for maximum reliable cooling is an easy and powerful metric that establishes the need to take action to resolve cooling issues. This data can be gathered with a hand held infrared temperature meter, with wired or wireless thermometers, or even taken from the IT device data stream. The energy used in cooling data centers is the largest component in a typical data center after the IT energy consumption. Research has found that the cooling energy can be dramatically reduced by properly managing the cooling air flow.

The primary opportunity to reduce energy consumption in a data center lies in the area of IT utilization. Some management teams measure this, most do not. If a company is serious about reducing energy consumption at a data center, looking into IT consumption is essential. Utilizations of 5-10% are common. Often the energy consumption at a data center can be reduced by 15% just be removing IT devices that were replaced in a technology refresh, but were never unplugged.

Data Center Metrics: Not So Confusing

- Tate Cantrell, chief technology officer of Verne Global (www.verneglobal.com), says:

The Green Grid (www.thegreengrid.org) has said in the past that the point for measurement of the IT Equipment Power is at the distribution point upstream of the computer equipment at the computer room power distribution units (PDU’s). For most companies, this means at the output circuit breaker at the upstream breaker panel. Google (http://www.google.com/corporate/green/datacenters/measuring.html) on the other hand has encouraged the industry to take the power measurement at the input to the server, excluding even the power cords of the computers in the IT Equipment Power measurements.

Google goes further to encourage data center managers to choose the utility side of their substation when calculating the Total Facility Power for the PUE calculation. This is a bit more specific and a little more challenging than the Green Grid definition of at or near the facility’s Utility power meter.

The well-run enterprise has a goal to increase top line revenue while keeping costs in check and preferably reducing costs over time. The purpose of the data center is to improve worker productivity within the enterprise. With proper direction, improved worker productivity should improve top line revenues. By improving on metrics, data center managers can impact the enterprise by improving worker productivity and thereby top line revenues, while honing efficiency and delivering the data center solutions with reduced cost impact to the organizations.

Without metrics, a manager cannot monitor the ongoing performance of the data center operation. And without a well designed set of metrics that are customized to the infrastructure at hand, a data center manager is unable to properly predict the trends of a data center and will be unable to effectively time projects for capacity increase.

Data Center Metrics: Today's Focus

- Dr. Joe Polastre, CTO and Co-founder at Sentilla (www.sentilla.com), says:

Metrics in use today tend to be application performance metrics, not energy efficiency metrics. These include things like transactions/second and users served per system.

What are the right metrics for energy efficiency in the data center? Metrics from the mainframe days are coming back, such as MIPS/watt and bits/kWh. Supercomputers are still benchmarked using MIPS/Watt, but when we moved to distributed computing we lost a lot of these benchmarks. Organizations like spec.org are trying to bring them back, and they're a first crack at deriving the true energy efficiency of applications/systems in the data center.

Data center managers need to be educated about metrics because it is financially responsible, lowers a company's bottom line, and delays capital expense of new data center facilities due to increased application optimization.

Data Center Metrics

- Daniel Kharitonov, Senior Staff Engineer at Juniper Networks (www.juniper.net), says:

There is obviously no point to compare the efficiency of power plant to that of a server or an Ethernet switch. However, it might make sense to fine-tune the metric within the equipment class. Let me give you an example.

When comparing network fabric, the main criteria is the efficiency of throughput, measured in Watts/Gbps. However, it helps to know the topology of traffic within DC - full-mesh connectivity may require a different class of network interconnect compared to aggregation (north-south) topology. Of course, the comparison will only be valid within a given class of devices. Ideally, the chosen element-level metric should reflect the primary mode of use in a live application. For instance, if a server performs well under specPower test (with many Java in-memory transactions, but very limited I/O and network operations), such server may coincidentally be inefficient for data mining. Knowing this, an additional metric may be needed to evaluate energy performance. It really does not matter if such tests are be done by vendors, 3rd party labs or customers - the only prerequisites are to maintain the integrity of test conditions and produce repeatable results.

Energy efficiency metrics add new criteria to supplier chain control. Ineffective products can be eliminated from the RFP altogether, or force the vendors to renegotiate contract terms. In both cases, the enterprise wins.

There are two main aspects to energy efficiency - operational cost and business image. Everyone can see a light bulb in a room, and many people have a habit to turn lights off at the end of their day. However, even small data centers are really "power lakes" compared to a lightbulb - which shows on electric bills. Improved operational efficiencies are the material savings that come at a relatively low cost - once in place, a datacenter stays up 24x7 for several years.

Another interesting aspect is how your company looks to the world. At this time and age, no one (not even the US army) can brag about careless spending of non-renewable resources. The world is getting smaller and many people realize we are in a same boat. This is why public, clear and well-thought energy strategy not only positions a company for long-term benefits, but also (indirectly) wins trust of clients and their families.

If someone reads an article about energy efficiency and thinks the authors know by heart what they talking about, this feeling can be projected to business relations.

The topic of energy effciency can be polarizing - some think it is very simple, while others say it is not sufficiently important, mature or standardized. But if you could read the article to this very point, you are already qualified to make powerful and informed decisions on what the future of IT is going to be. The time to act is now.