David J. Cappuccio
Research VP
6 years at Gartner
41 years IT industry
David J. Cappuccio is a managing vice president and chief of research for the Infrastructure teams with Gartner, responsible for research in data center futures, servers, power/cooling, green IT, enterprise management and IT operations. Read Full Bio
by Dave Cappuccio | November 9, 2012 | 1 Comment
Data Centers have gotten a bad rap of late, with both the press and senior executives putting pressure on them to improve overall efficiency and reduce operating costs (yet again). The focus for the last few years has been around energy efficiency and the PUE (Power Usage Effectiveness) metric developed by The Green Grid. PUE is a great metric, and when used wisely it can help organizations easily improve energy efficiency oftentimes by 20% and more. Unfortunately for Data Center managers, the efficiency gains attained with a PUE focus were almost all on the Facilities side of the equation, and while IT and the Data Center may have gained some benefits (e.g. improved cooling), the lion share of the operational savings was applied to the Facilities budget (unless of course you’re one of those rare companies that gives IT it’s own power budget).
But realistically the headlong rush to better PUE’s has done little to improve data center efficiency. Having a facility with a great PUE is one thing, but if my data center is highly underutilized, or if the resources are poorly managed, IT has not solved the real problem of trying to get the most out of the resources we already have. The other problem with PUE is that as Data Center managers strive for more energy efficient IT equipment, they could inadvertently degrade that wonderful PUE the Facilities team reported last year. Take a hypothetical data center with an average PUE of 1.5. If IT decides it’s time to do a technology refresh on some servers, and bring in the current generation as replacements, the overall performance and productivity of applications will increase, but because of the energy efficiency improvements vendors have made, the overall power draw for IT could very easily go down. When that happens the ration of total building power to IT load gets worse – negatively impacting PUE. So a great decision by IT could easily create a bad impression of Facilities, unless everyone understands the overall value of what was done.
So given that rather long preamble, I’ve been thinking about taking the same concept of PUE (optimum vs. actual usage) and applying it to the Data Center proper in order to create a resource efficiency metric. The problem with creating a metric like this is that all Data Centers are not created equal – and don’t have the same type of equipment or configurations. So given that caveat, I’d like to propose creating a metric around the most common resource available in most Data Centers – the Rack Unit, or RU. A standard rack today has 42U, others have 48U, 50U and more, but the single RU itself is something we can track.
So here is the basic idea – and I’ll be writing more on the RUE and RUiE metrics in Gartner’s published research. Let’s assume the following just for illustrative purposes:
300 Rack maximum capacity (approximately 9,000 feet of floor space).
Standard 42U size
180 racks are currently installed, and average 65% utilization.
The maximum RU count at capacity is 12,600 (300*42)
The Installed RU count is 7,560 (180*42)
The utilized RU count is 4,914 (7,560*65%)
Using the same construct as PUE, we take the maximum and divide by the actual (12,600 / 4,914) and come up with a ratio of 2.56 (where capacity would be 1.0)
The Data Center RUE is now 2.56 and can be track fairly easily to monitor both growth and efficiency.
Using the reciprocal (1/RUE) yields your utilization; RUiE = 1 / 2.56 = 39%.
Now the big flaw here is an obvious one – nobody wants to get to perfection. An RUE of 1.0 would indicate you were completely out of room – and that’s not a metric I’d want to attain. However, using this same idea you could modify the maximum capacity to an optimal capacity. Lets assume no rack should exceed 90% capacity as a target. The results would look like this:
The maximum RU count at capacity is 11,340 (300*42*.90)
The Installed RU count is 7,560 (180*42)
The utilized RU count is 4,914 (7,560*65%)
RUE now becomes 2.31 (11,340 / 4,914) and the RUiE is 43% ( 1 / 2.31).
Still a usable metric, but one with where if I reached my optimal goal of 1.0 I’d still have some space left while I built (or moved to) the next Data Center.
Food for thought? Comments welcome.
Category: Data Centers Food for Thought Tags: Data Center Efficiency, Metrics
by Dave Cappuccio | June 7, 2012 | Comments Off
When faced with planning a new data center the question of how much space will be needed is potentially the most difficult to determine. That said, the answer is often one of the quickest made – with the least amount of analysis, and I would suggest that it is rarely correct, and in most cases the final size is far larger than what is actually needed.
The first mistake many people make is to base their estimates on what they currently have – extrapolating out future space needs according to historical growth patterns. It sounds like a logical approach, but there are two fundamental problems; the first being an assumption that the floor space currently used is being used properly, and the second is a 2 dimensional view, or the assumption that usable space is a horizontal construct, rather than a combination of both horizontal and vertical space.
Many times I have seen Data Center managers or Facilities teams start with the following assumption: We are out of (or near) capacity in our data center, therefore when we build next we will need more space. If we have 5,000 feet today we must need at least 7,500 or more to sustain our growth. The error is that the focus is on square footage, not compute capacity per square footage.
By looking at compute capacity as the metric things begin to change rather quickly. As an example, lets take a typical environment of 40 server racks. In a high percentage of data centers today these racks would be populated with servers 1 or 2 generations old, depending on corporate refresh cycles, and the average server would be a standard 2u height. The racks would rarely be nearing physical capacity but might actually be maxed out in logical capacity due to power or cooling constraints at the rack level (the mantra to avoid creating hot spots in data centers has actually made floor and rack use a lot less efficient).
Given a 60% load capacity on average (again, to avoid hot spots), our example would yield an average of 13 physical servers per rack (assume 42u racks) and 520 physical servers. Given 30 square feet per rack (which includes aisle ways, door swing space, etc), the 40 racks would require 1,200 feet of floor space.
So how big should the next data center be? If we assume 15% CAGR as an average growth target, in 10 years our small IT room would need to support at least 160 racks with over 2,000 physical servers, and would require almost 5,000 square feet of floor space.
But – what if we thought both vertically and horizontally? The above all assumes things stay status quo and I acquire the same type of equipment and apply the same configuration policies throughout. But lets assume whatever floor size you design was created to allow full use of rack space without the fear of hot spots (and there are many ways to do this with a great deal of expense). Taking the same 40 racks, if pushed to 90% capacity on average (leaving some room for switches, etc), and upgrading the existing server base over the next 2 years to 1U servers would support 1,520 physical servers.
So a data center of the exact same size, containing 40 racks, with the proper design, would support 15% growth every year for at least 8 more years. Now the question becomes – do we build it bigger to support the original target of 2,000 servers, or will a future technology refresh within the next 8 years double our capacity yet again?
Doing some simple spreadsheet exercises and asking these “what if” questions can yield some startling results when it comes to capacity estimates. And the logic works with servers as well as storage, as each device category continues to decrease in size, improve in capacity and performance, and reduce it’s power consumption per unit of work with each new generation.
If we were to look at these performance and density trends and make the assumption that the curve will continue – even at a much slower pace, it becomes clear that even small data center environments can have significant growth rates (well in excess of 20% CAGR), while maintaining the exact same footprint over the next 15 to 20 years.
Food for thought – and as an aside – food for thought when contemplating the life cycle of a Container based data center as well.
Category: Consolidation Data Center Design Data Centers Food for Thought Tags: Data Center
by Dave Cappuccio | April 6, 2012 | Comments Off
Cooling with the data center has become our achilles heal in many cases. Historically the folks in IT had relatively nothing to do with heat or cooling management, this was strictly under the purview of the facilities team (after all, if it wasn’t IT gear, it didn’t count). In todays world though the IT team has to get involved, since they are the ones that need to live with (and fix) the problem.
In this series of posts I’ll posit 10 of the easy steps you can take to solve, or mitigate the cooling issue at your site.
9. Technology Refresh
Using technology refresh as a cooling solution may seem counterintuitive to many people, but it in fact is a proven solution for many. An interesting trend has been under way that could help IT organizations solve multiple problems simultaneously (see “Grow Disk Storage 800% or More, Without Increasing Power or Cooling Costs, in the Same Space”). The problems are intertwined in almost all data centers: capacity, space and power. Each issue is impacted almost every time equipment is added or changed on the data center floor. Historically, capacity planners focused on new application growth and a continuous drive toward virtualization, while keeping existing equipment in “maintenance mode,” trying to get the most work out of the equipment over the longest time possible — especially when faced with tight budgets. It turns out that this seemingly prudent use of resources was not, or will not necessarily be, the most prudent thing to do. One reason is that the energy requirements of older servers, in some cases, is three to four times greater than new equipment.
In recent years, x86 server performance has been doubling (or increasing even more than that) with each new generation, while at the same time becoming more energy efficient. Doubling the performance and halving the power in the same space is a sound cost-saving concept. When you look at AMD or Intel performance numbers over the last few generations, and then compare the energy consumption for each of those generations, you’ll notice a dramatic increase in processor performance, while at the same time seeing a significant reduction in energy consumption (and heat generation) for those servers. An equipment replacement policy that is escalated (rather than deferred due to capital constraints) can in fact have the added benefit of reducing energy (and cooling) requirements while also reducing physical capacity (smaller footprint) and increasing performance.
10. External Augmentation
For data centers nearing capacity, either of physical floor space or the facilities infrastructure to support the IT load, the idea of external augmentation is beginning to resonate. In actuality this technique is not about augmenting an existing environment, but about offloading some percentage of workload elsewhere in order to free up power, cooling and floor space in the existing data center for future growth. Depending on the age and location of the data center the type of workload involved can vary greatly. In some cases data centers are so old there is a great risk of impacting business outcomes with an extended outage and therefore the high risk, mission critical systems are potential move candidates while improvements or a retrofit project is completed.
In other cases the data center may in fact be very robust and highly fault resilient, but cannot handle the current growth trends. In these cases offloading non-critical work (e.g. back office systems, test/development) may become a viable alternative to building out a complete new data center. In either case the offloading is often considered a short term (e.g. 2 year) solution while the optimal solution is developed.
Category: Data Centers Power and Cooling Tags: Cooling, Data Center Efficiency, Energy
by Dave Cappuccio | March 30, 2012 | Comments Off
Cooling with the data center has become our achilles heal in many cases. Historically the folks in IT had relatively nothing to do with heat or cooling management, this was strictly under the purview of the facilities team (after all, if it wasn’t IT gear, it didn’t count). In todays world though the IT team has to get involved, since they are the ones that need to live with (and fix) the problem.
Well the good news is that in most older data centers (older being 10+ years), there are plenty of low hanging fruit to choose from when deciding what project to undertake in order to develop a more efficient cooling environment within the data center.
In this series of posts I’ll posit 10 of the easy steps you can take to solve, or mitigate the cooling issue at your site.
7. Shut down CRACs
It is often the case that a data center can have too much cooling rather than too little. Many companies find themselves with a sizeable data center that is cooled to a consistently low temperature across the floor space, even when some of that floor space is either empty, or contains equipment that needs minimal cooling. In these situations, especially within older data centers, the solution can be as easy as physically shutting down some CRACs. This simple technique is often overlooked by IT for a couple of reasons. First, the responsibility for Infrastructure equipment like CRACs falls under the purview of the Facilities team, and therefore IT staff rarely think about CRAC efficiency. The second reason, especially on older equipment, is that the system has been set to a standard fan speed (often High), and left in that condition as a standard operating procedure. These fans in many cases are the most energy hungry devices on the data center floor, so any opportunity to either moderate them (item 2) or shut them down should be taken advantage of.
8. Shrink Floor-space
Companies that have experienced M&A activity, or those that have employed newer (smaller) server and storage technologies often find themselves with more floor space than is actually needed. In many cases IT looks at this space as a value-add, as it provides room for potential growth in years to come. However, this excess space also needs to be conditioned and is often kept at the same temperature as the rest of the floor since it’s all one contiguous space. In the past few years we have seen an increasing trend to shut down this space, freeing it up for other uses. By walling off excess space IT can reduce monthly operating costs (reduced energy use), while at the same time freeing up possible office space, IT work areas, or even releasing leased space. In situations where the asset is owned and IT isn’t quite sure how much growth they can expect over time, the use of temporary moveable walls might be a viable alternative. In either method the objective is to reduce the conditioned IT space down to what is absolutely needed for the next few years, and not keeping all available space just because it’s there.
Category: Food for Thought Tags:
by Dave Cappuccio | March 28, 2012 | Comments Off
Cooling with the data center has become our achilles heal in many cases. Historically the folks in IT had relatively nothing to do with heat or cooling management, this was strictly under the purview of the facilities team (after all, if it wasn’t IT gear, it didn’t count). In todays world though the IT team has to get involved, since they are the ones that need to live with (and fix) the problem.
Well the good news is that in most older data centers (older being 10+ years), there are plenty of low hanging fruit to choose from when deciding what project to undertake in order to develop a more efficient cooling environment within the data center.
In this series of posts I’ll posit 10 of the easy steps you can take to solve, or mitigate the cooling issue at your site.
5. Airflow
The primary force for cooling in data centers is air, and the control of airflow can be a simple method of increasing cooling efficiencies with minimal expense. In many cases when equipment is installed in racks and the rack has open space, server administrators fail to install blanking panels (or defer it until they have time – which often never happens). By not installing these panels hot air from one server easily moves up the rack, contaminating (heating up) equipment above it. These panels are design to control this flow and should be used whenever possible.
A second and more basic method of improving airflow is the remove any blockages from under the floor itself. The accumulation of power and data cables over the years, especially in older data centers, is often when of the biggest impediments of good airflow, and can restrict efficient cooling by as much as 30%. The issue is often that due to the multiple layers of cables under the floor, pulling them out can be a risky issue for data centers that are running production workloads, as mislabeled cables may get pulled which then disrupt active systems. These projects are often best managed as weekend/holiday endeavors, often after a significant amount of planning.
Additionally there are air flow systems available for redirecting air flow under floors via fans and sensors. Example vendors; Tate Floors, Triad, Legrand.
6. Chimneys
The issue with hot racks is that hot air generated within the rack can leak out into the floor space and be reintroduced into another rack, thus aggravating that racks cooling process. A simple solution developed over the past few years is to create a chimney above the rack that redirects the hot air directly upwards towards the plenum for removal from the data center. These chimneys have been built by Facilities teams, or can be acquired by specialty vendors, and while not the most attractive device in the data center, they do what’s necessary to improve cooling efficiency at a very low cost. Typical vendors; Chatsworth, HP, Great Lakes)
- Posted using BlogPress from my iPad
Category: Food for Thought Tags:
by Dave Cappuccio | March 25, 2012 | Comments Off
Cooling with the data center has become our achilles heal in many cases. Historically the folks in IT had relatively nothing to do with heat or cooling management, this was strictly under the purview of the facilities team (after all, if it wasn’t IT gear, it didn’t count). In todays world though the IT team has to get involved, since they are the ones that need to live with (and fix) the problem.
Well the good news is that in most older data centers (older being 10+ years), there are plenty of low hanging fruit to choose from when deciding what project to undertake in order to develop a more efficient cooling environment within the data center.
In this series of posts I’ll posit 10 of the easy steps you can take to solve, or mitigate the cooling issue at your site.
3. Ambient Temperature
Ambient temperature, or the average temperature in the data center, is often the easiest efficiency target, and often the one that is most often overlooked. Historically data centers were operated with room temperatures in the 68°F to 71°F range, primarily due to concerns about overheated IT equipment. These concerns were first developed by data centers during the mainframe era and have been carried forward to all data centers, regardless of equipment mix. However, todays server, storage and networking equipment have operating temperature variances that can easily exceed 95°F and still remain within the manufacturers guidelines.
Now we do not advocate running data centers at those high temperatures, but ASHRAE (American Society of Heating, Refrigeration and Air Conditioning Engineers) has published it’s 2011 guidelines for data centers which recommend average ambient temperatures between 74°F and 80°F. Raising these temperatures can be one of the fastest ways to save operating costs there is, as studies have shown that raising the ambient temperature by 1°F can save upwards of 3% in energy costs. To obtain significant energy reduction while still maintaining a comfortable working environment Gartner recommends that data center operators consider an average ambient temperature of 78°F.
4. Hot and Cold Containment
The placement of racks within data centers has changed over the years, primarily due to an ever-increasing need to manage the heat exiting them. For the past 10 years or so the idea of hot and cold containment has become standard operating procedure (see 1 above). While this did not completely solve the heat problem, it did position the racks such that the hot air leaving one row would not immediately be drawn into the next row. Server administrators were still limited to how many devices to put into each rack though, as higher density racks would often create hot spots on the floor. These hot spots were then controlled by placing higher density racks across the floor space, essentially sharing their higher air temperature outputs with the rest of the floor.
Well designed data centers today have solved the hot spot issue by concentrating high density racks into hot or cold containment isles, rather than by spreading them around the floor. These aisles are designed so that all the heat leaving the racks contained within the row via walls, and then is quickly channeled upwards to the plenum, thus insuring hot air is “shared” with the rest of the data center. These hot containment zones can be constructed via sheet rock, heavy-duty plastic sheeting, or self contained units pre-designed by vendors. Hot or cold containment zones have two distinct benefits; the reduction of energy required to cool the data center floor through the elimination of heat leakage from the ends of rows, and the ability to fully utilize rack densities, allowing increased kilowatts per rack in the containment zones, thus increasing the usable rack space.
Category: Data Center Design Data Centers Power and Cooling Tags: Cooling, Design, Green IT
by Dave Cappuccio | March 24, 2012 | Comments Off
Data Centers have been an integral part of the enterprise for almost 50 years, but the question of whether we should build more of them, or just let someone else manage them, has been with us since the beginning. And has yet to be resolved. Whenever a data center build project is initiated it seems the first questions that need to be resolved are whether to build, and spend a significant amount of capital in the process, or colocate, thus minimizing the capital spend, but giving up control.
And the answer of course, sad to say, is “it depends”. For those with fairly static workloads – those that are steady state and run at consistent, predictable levels day in, day out, the decision is often an easy one. Colocation provides almost all the benefits of a buildout – a state of the art building, often Tier 3+, reliable infrastructure, more than adequate power and cooling for the foreseeable future, and almost immediate availability. Growth outside of the contracted environment is not an issue for these environments since the applications are predictable.
For enterprises with very active application portfolios, or who experience unplanned M&A activity the colocation decision is a bit more problematic. Identifying the resources needed today, and even projecting needs a year or more out is fairly easy, but newer applications, unplanned business opportunities, or M&A activity can cause dramatic changes in the IT infrastructure needs. For these environments colocation becomes an issue because of the unplanned growth – and the realization that the likelihood of additional space being available within the same facility in a few years is probably negligible, simeply because in order for the provider to maximize their profits they need to lease out the buildings space as fast as possible. It’s possible to lock in growth space, either adjacent to your space, or elsewhere in the building, but at a cost up front – essentially forcing you to pay for unused space. An alternative could be contracting for a right of first refusal clause whereby you would have the option of lock in the last space before it was leased to someone else, but again this would entail paying for unused space well before it was needed.
Another alternative is beginning to be used by enterprises, that of looking at colocation as augmentation space for existing data centers, rather than replacement space. This is evolving in three different forms as follows:
1. Temporary Space. Organizations that find themselves nearing capacity of their data centers but do not have the time, or budget, to build out their final solution right now are using colocation as a temporary stop-gap. Space is leased for a short term contract, perhaps 2 or 3 years, and is used to absorb growth while the new data center is designed and built. The downside of this is that eventually 2 moves are required – first to the colo facility, and then to the final destination, but if planned properly this can cause limited risk to the business. The upside is that it gives an enterprise breathing room to properly plan, design and execute a data center build project.
2. Non-Critical Space: When a robust production data center is nearing capacity a design/build project is a risky business. Not only does a new site need to be prepared with high levels of redundancy, scalability and availability, but the physical move is fraught with risk to day to day operations. Done right these projects take 3 to 4 years and the process of the move is often many times more complex than the construction project itself. In these situations colocation becomes an adjunct to existing data center space – for non critical systems. By moving non-critical or commodity systems (e.g. back office, test/development) to a colocation facility valuable floorspace, power and cooling resources are freed up for future production growth. This is a low risk scenario since the moves are for non-business critical only, and it becomes an operating expense rather than capital expense impact only. Also, due to the non-critical nature of the applications the colocation provider only needs to support a strong Tier 1 or Tier 2 environment, thus lowering the monthly costs.
3. Mission Critical Space: Many data centers today are artifacts of the past and were not designed to support todays high performance systems. These sites were often designed as server rooms or grew organically , adding space from within the building in reaction to growth (vs. planned expansion). As such they either can’t support the power and cooling needed for todays rack based or blade systems, or have inadequate redundancy in the mechanical infrastructure, or both. In these situations companies often are faced with a choice to rebuild/retrofit – which entails significant business risk, or build a new site at a large capital expense. With the current economy the retrofit option has been forced on many IT shops due to reduced budgets and in these instances colocation is becoming a viable short term alternative for production workloads. Moving mission critical work to a more robust colocation facility solves the problem of infrastructure availability and redundancy – and reduces risk to the business, which in turn frees up floor space in the older facility which can then be used as the beginning of a retrofit project. Since the retrofit will not impact non-critical systems it has reduced risk, can be extended over a longer term to spread the capital spend over multiple years. Once completed, a migration of critical systems back to the data center can be accomplished.
The are other variations on this theme, especially when considering business continuity and disaster recovery planning, but the bottom line is that looking at colocation vs. build as an either or decision is no longer a viable way to view your options.
- Posted using BlogPress from my iPad
Category: Food for Thought Tags:
by Dave Cappuccio | March 23, 2012 | 1 Comment
Cooling with the data center has become our achilles heal in many cases. Historically the folks in IT had relatively nothing to do with heat or cooling management, this was strictly under the purview of the facilities team (after all, if it wasn’t IT gear, it didn’t count). In todays world though the IT team has to get involved, since they are the ones that need to live with (and fix) the problem.
Well the good news is that in most older data centers (older being 10+ years), there are plenty of low hanging fruit to choose from when deciding what project to undertake in order to develop a more efficient cooling environment within the data center.
Over the next few posts I’ll posit 10 of the easy steps you can take to solve, or mitigate the cooling issue at your site.
1. Hot Aisle/Cold Aisle
As one of the oldest techniques in data center efficiency, placing rows of racks back to back and front to front is a simple method to redirect some of the hot air exiting the back of your racks into a common area. Without this configuration the hot air will exit the rack and be pulled into the front of the adjoining row, thus creating hot spots and very poor distribution of cool air. Creating hot and cold aisle configurations can improve air flow and cooling efficiency by as much as 10%.
2. Variable Speed Drives – Variable Frequency Drives
One of the largest consumers of electricity in data centers are the air handlers and/or air conditioners (CRAH’s and CRAC’s). These units have very powerful fans which need to move large amounts of air across the data center floor to the IT equipment. On many older units the fans spin at a continuous rate and the airflow remains continuous, regardless of how much is actually needed (e.g. a mostly empty data center would get the same airflow as a nearly full data center). Unless IT or Facilities teams are diligent, these fans are normally left on one speed all the time, creating a very inefficient environment. Newer CRAHs and CRACs have variable speed drives which can automatically sense temperature variations on the floor and therefor modulate the airflow, and fan speed, accordingly. Sophisticated units allow the synchronization of many units across a data center so the optimal fan speed, air flow, and energy consumption can be achieved. Implementing variable speed drives, synchronized, can reduce energy consumption by up to 15%. Example vendors; Emerson, Schneider Electric, Stulz.
Category: Data Center Design Data Centers Food for Thought Green IT Power and Cooling Tags: Cooling, Data Center, Design
by Dave Cappuccio | May 20, 2011 | Comments Off
I would be remiss if I didn’t mention the Infrastructure Operations and Management conference coming up the 2nd week of June in Orlando. At a time when every IT dollar counts, our 5-track agenda provides the strategic guidance and pragmatic recommendations needed to demonstrate the business value of I&O as you meet the challenges of modernization, transformation and convergence. We’ll extensively explore today’s I&O hot spots, including virtualization environments, assessing a public vs. private cloud strategy, aggressively optimizing legacy systems, achieving operational excellence despite limited resources and much more. Get ready to identify new business models, data center design strategies and technology approaches that can deliver ROI fast, efficiently and effectively. Your takeaway: the ability to leverage real-world experience and knowledge gleaned from more than 40 sessions, including analyst-user roundtables and end-user case studies.
The agenda will feature comprehensive tracks to drill down on the I&O topics most important today, with track sessions containing 6 to 8 highly focused presentations from Gartner analysts. .
If you’re going to attend, stop me in the hallway and say hello.
Category: Data Centers IT Operations Tags:
by Dave Cappuccio | May 19, 2011 | 1 Comment
Over the past few years, data center design has shifted away from the traditional idea of building out large expanses of IT-ready floor space to an era where smaller component-based solutions are the trend. A variety of different options from vendors have been introduced, and the two prevailing trends we see are container-based prebuilt solutions or modular component-based assembled on site) solutions. Since these solutions are fairly new to the market, and the terminology can be confusing, this research will outline the differences between the solutions and offer possible use case scenarios for you to consider.
What Are Containers or Container Based Data Centers?
A data center container is a shipping container set up to accommodate IT equipment. The typical data center container is based on ISO standards for ease of shipping; also, it may be modified in a number of ways to better support the secure and practical use of IT equipment. The basic equipment that most of these containers are designed to support includes servers, storage and networking gear. In addition, containers may be designed to support some combination of uninterruptible power supply (UPS), generators and/or chillers, with some of that equipment supported in the same containers as the servers and storage equipment, or in separate and distinct containers. Generally, data center containers have some connectivity elements, so that power input, cooling capabilities (e.g., water pipes) and network traffic can be fed into the container from the outside.
Containers are designed to be weather resistant, and, in some cases, are weather hardened for use in extreme environments, although most use cases see them being implemented within existing buildings or shells of buildings. Containers are also designed to be TIA-942 Tier 3 capable (depending on supporting infrastructure), and focus on high levels of energy efficiency, typically with a PUE of 1.3 or below.
What Are Modular or Component-Based Data Centers?
Modular data centers evolved from the basic premise of containers — that, if designed appropriately, extreme levels of performance could be attained in data centers using a consistent design technique, and capital costs could be reduced by standardizing components, construction and the supply chain. The modular design approach has actually split in two directions over the past few years. While some vendors have focused on the overall design of a complete data center solutions (e.g., HP FlexDC and i/o’s Data Systems i/o Anywhere), others have moved the modular design concept down to the rack or row level (e.g., APC InfraStruxure or Emerson SmartAisle).
In modular data centers, a core infrastructure design is developed that will eventually support multiple IT areas. The IT areas can be added as needed to support future growth, until the base power requirements are exceeded, at which time either the base module can be upgraded or another modular unit is put into place. Most of the vendor offerings in this space start with a base module of 250kVA or more and IT space of at least 500 to 2,500 square feet, which can then be scaled, but, as the market evolves, we expect to see a greater variety of choices (essentially modules on demand) to support individual needs.
In modular component designs, vendors have begun to focus on self-contained row or rack solutions where a predefined unit (e.g., eight racks) is delivered on-site and is already configured for power and cooling support. These units can be as simple as basic racks with monitoring and directed air cooling, or as complex as completely self-contained environments with either air, water or refrigerant-based cooling for extreme densities. UPS support can be delivered by the customer or integrated as part of the solution.
Bottom Line
When planning for data center growth, it is important that all alternatives be reviewed. Newer modular design techniques and container-based solutions should be a critical piece of your analysis. When used appropriately, they can solve specific problems, while reducing capital costs and the time it takes to implement new capacity.
Category: Data Center Design Data Centers Green IT Power and Cooling Tags: Container, Data Center, Modular
