When the world watched as Hurricane Sandy made landfall, everyone’s thoughts were with those living in New York and New Jersey. Everyone was just hoping people could remain safe. The superstorm devastated many locations and tragically had casualties. In the aftermath of the storm, people had to pick up the pieces of the devastation and among the problems of Superstorm Sandy were the issues that many data centers had during the storm. In a world that moves very fast and constant functionality is critical, many data centers experienced extended and very costly periods of downtime. Many data centers did not have proper disaster recovery plans in place. Unfortunately, natural disasters, regardless of your location, are not something you can anticipate fully or predict so the best thing to do is to try to plan for any potential contingency and have a well designed plan that all employees are prepared to implement. Everyone needs to know what the plan is and what their role and responsibility is should a disaster occur. There are some important lessons that we can learn from Hurricane Sandy so that our data centers can be better prepared in the case of an emergency.
The important lesson to learn is that building infrastructure and power capabilities are of critical importance. If you are given any amount of time to prepare before a storm or event occurs, checking to make sure backup generators, equipment and structure are fully functioning is very important. By doing this, should any last minute adjustment need to occur, you will have the opportunity to ensure your facility is prepared. If you have the opportunity to design your data center from the ground up, considering natural disasters and major storms should be part of the design process for the infrastructure. Things like location should be considered when trying to avoid things like water damage. The basement may not be the best location for a data center or any critical electrical or mechanical assets as they could become gravely damaged. Backup generators are incredibly important and should be routinely checked. They could mean the difference between an extended and expensive downtime. Additionally, many data centers found it difficult to obtain fuel for their generators in the wake of the storm. Some other options, diesel or natural gas generators, should be considered as an alternative.
Additionally, there are other options to consider when preparing for unexpected events. You may want to consider having an additional location where data is backed up that would theoretically not be affected by the same storm. You may also want to consider implementing the cloud, a much hyped technology. According to Zeus Kerravala of ZK Research, “One of the most understated use cases for the cloud is disaster recovery. The cloud is built for backup and recovery, with geographically disbursed data centers.” These are just a few additional preparedness options to consider.
While we cannot predict where or when the next natural disaster like Hurricane Sandy will occur, we can certainly learn lessons from the superstorm for our data centers. By having disaster recovery plans in place that are thorough and prepared for any contingency, we can avoid expensive downtime and keep our business up and running.
Posted in Back-up Power Industry, Computer Room Design, Data Center Build, Data Center Design, data center equipment, data center maintenance, Uninterruptible Power Supply, UPS Maintenance
Tagged Data Center, Data Center Construction, data center design, data center equipment, data center maintenance, Uninterruptible Power Supply
With the economy slowly in recovery, we are all still trying to save money. Running a data center requires a lot of energy usage as well as other expenses. Every year we learn more and more about energy efficiency, “going green,” and best practices for data centers that can help lower operating costs and save money. There is also certainly the added benefit of being more environmentally friendly, which we can all appreciate. In a data center that focuses on having one hundred percent uptime if possible, things like uninterruptable power supply systems, computer room air conditioners, cooling and humidity control systems, and more can be very costly. Many attempts to go green with your data center or save money may seem daunting and expensive themselves so some are reluctant to take them on. But, you do not need to innovatively break ground with green initiatives in order to become more energy efficient and save money. Companies like Google and Facebook make headlines for their creative efforts in energy efficiency. It has recently been reported that Google’s data centers use recycled waste water to help cool their data centers, essentially, flushing a toilet helps cool their data centers. It is incredibly creative and great that large companies can take these steps to become more energy efficient but there are baby steps that many companies can implement as well.
To begin making changes and help your data center save money you will need to begin with knowledge. You need to learn where energy expenditures are going and look at the trends of your data center to see what is using the most energy. By starting there you can make an educated decision as to where to begin with your energy efficiency changes. It may not be worth your time and money to make the change to energy efficient lighting first if it is only less than 1% of your overall energy expenditure. You want to start with what is the most doable and will save you the most money immediately. You also need to evaluate the design of your data center and its infrastructure. The placement of your computer room air conditioner exhaust or intake may fool it into working harder than it needs to and will cost you more money. Consulting a data center design company is a great way to become informed of what could be problematic for your data center and if you should redesign your data center. A professional opinion will give you the insight into potential energy drains that you make have overlooked. Something as simple as turning off lights and unused equipment can save thousands of dollars! Also, when checking your data center design you will want to make sure you have optimal air flow as poor air flow can lead to vastly increased cooling expenditures.
The key to saving money in your data center is taking it step by step. It does not have to be overwhelming or incredibly expensive to make your data center more energy efficient. Start with consulting data center design professionals who can guide you towards the most savings for your data center.
Data centers utilize a lot of equipment that also uses a lot of energy. The amount of energy used emits a lot of heat. If overheating occurs, it can cause a lot of problems including potential fire hazards and the possibility that equipment could become overheated and fail. Data centers cannot afford to have the downtime that these problems would cause. That is why sufficient cooling is needed in any data center. It is crucial that, not only is the area cooled but, that heat is actually removed from the space. Every data center is different and has it’s own unique cooling needs. Meeting a data center’s cooling needs should be a top priority.
A great place to begin is with the hot aisle/cold aisle technique. It is a simple yet practical and efficient way to contain hot air and use row-based cooling. By preventing the mixture of hot air and cold air you can focus your cooling where it is most needed. Using this technique provides you the ability to effectively cool your data center while reducing your overall energy consumption and also will save money on your energy bills.
Next, a great way to sufficiently cool your data center is with rack based cooling. You can ensure that the cooling is delivered directly to the rack, keeping it, and it’s components, at an optimal temperature. Another type of cooling used in data centers is called CRAC (computer room air conditioners). They are very commonly used in data centers. These nits monitor the data center’s temperature, air and humidity. They work differently than traditional air conditioners because they cool but also remove energy from the air and redistribute it to an outdoor space.
In addition, something that has been used for a long time is a raised floor. Some debate the merit of raised floors now with new technologies available. Raised floors allow cool air to flow under the floor and into the server area through specially designed perforated tiles. An alternative to raised floors is a slab floor. With a slab floor air conditioning systems are installed above racks and forced down towards the floor. There has been much debate within the data center industry as to which technique is better. The issue is a complicated one and one that should be decided on an individual basis, based on the needs of each unique data center.
There are, of course, more specific things to consider for cooling a data center such as glycol-cooled chillers vs. air-cooled chillers and more. If you begin with these basics of data center cooling you will have the perfect jumping off point for your data center. A professional data center designed can guide you in the right direction based on your data center’s individual needs. Proper cooling of your data center is crucial to the efficient functioning of your equipment. It will help your equipment last longer and function efficiently.
When you have spent countless hours and a lot of money designing a data center, the last thing you want is to have a problem and lose precious data without a proper backup system in place. The loss of this data could take enormous amounts of time to recover and you may never fully recover all of it. It would be an incredibly expensive mistake! There are many things to consider such as how to backup your data, how often to backup your data, do you have enough space to backup your data, and more. By choosing to implement an effective backup system for your data, you can save yourself a big headache, hours of time, and incalculable amounts of money. Whatever system you choose to implement, it is imperative that you periodically check your backup system to ensure that it is working correctly and that you can in fact, access the data, should a problem occur. Having a peace of mind that you have a well functioning backup strategy in place for your data is invaluable.
Backup for a data center is not as much about the product or software you use, as much as the process and strategy you implement. One way in which you can properly protect your data center is to implement redundant, duplicate or multiple different forms of backup. Do not pull all of your eggs in one basket, so to say. Relying on one backup system could be detrimental. You should also plan the frequency with which you backup around your business needs.
As with any system you put in place for your business or data center, you should put your backup strategy in writing. When you have your strategy in writing, if a problem occurs, anyone who is authorized will know exactly what to do and everyone will be on the same page. Update your written strategy as any changes or adjustments to the plan occur. Write your strategy in easy to understand ways so that there will be no confusion should the need for recovery be rushed or should the data center be in crisis mode. Along with having a written backup strategy, you should make sure you have properly trained everyone as to their role in the data recovery process. The last thing you need when trying to recover your important data is to have people scrambling and confused.
There are many different specific ways to backup your data such as disk, cloud based, online, tape, and more. You can choose specific types of backup based on your data center needs. The important thing is that you consult a professional that can help determine your needs and the appropriate way to backup your data. With the recent superstorm Sandy affecting many businesses and data centers, it is an important reminder of how crucial backup is. According to Buzz, “With 90% of people not backing up data, this means much of that data is gone forever. For businesses, some will shut down due to this loss of data.” 93% of companies that lost their data center for 10 days or more went bankrupt within a year, according to the National Archives and Records Administration. Data loss is often incredibly personal, but it also has a tremendous economic impact. We cannot predict what tomorrow will bring, but, with proper backup strategies in place, we can protect our business.
When you invest in building a well designed data center you want to ensure that your data center will give your business security, have a superior overall service level, and be able to adapt to any needs that may arise in the future. There are, of course, all the technical aspects, like cables and wires that a data center needs but there is more to great data center design than that. With the extremely high costs of downtime, and the headaches that accompany it, you want to ensure that your data center will be able to remain up and functioning under any circumstances. The last thing you want to have happen is spend a lot of money and time designing your data center and find that in three or four years, your data center no longer sufficiently supports your business and is, in essence, obsolete. By tailoring a data center to your needs but also creating flexibility in your design you can have a data center that will continue to service your business for the long term.
The first data center must have may be unseen but is incredibly important: flexibility. You have to consider the future needs of your data center and that expansion and change will be inevitable at some point. You need to be able to grow within your existing data center. Things like adding extra racks and extra server space as well as the ability to add extra power to accommodate changes is crucial to saving time and money down the road. It can be our instinct to want to cut corners at every turn but this can lead to a lot of problems down the road for data centers. Flexibility and expandability is crucial to the design of any data center.
The next essential thing you will need in your data center is adequate and steady power. The reliability of power in a data center may be the most crucial thing. Without constant, reliable power, should any external problem occur that shuts down power in the area, your data center could experience extended and costly lengths of downtime. We cannot anticipate when a power outage will occur but we can prepare for the worst. By ensuring you have adequate power and employing the help of backup power, you can have pace of mind that your data center will be prepared. Backup power includes generators as well as uninterruptible power supply (UPS) that will kick in should a power outage be experienced.
Additionally, your data center will need adequate cooling to support your data centers changing needs. Data centers use a lot of energy and generate a lot of heat. You will need adequate air conditioning and may want to use techniques like the hot aisle/cold aisle technique that can help racks remain cool and properly functioning.
Lastly, it is vitally important that your data center have proper security. You cannot overlook the sinister aspects of this world. When data and crucial company information is involved, there is the potential that someone will try to tamper with or steal data. You need to have both proper physical security as well as virtual security.
If you begin with these data center must haves you will set yourself up for success. When a lot of money and time are invested in the design of a data center you cannot cut corners. By properly preparing for unforeseen circumstances and future needs your data center investment will be wise and will continue to pay off in the future.
Whether or not we live in the area, we have all heard about Superstorm Sandy and it’s devastating effects on the East Coast. The hurricane turned superstorm ravaged areas of New Jersey, New York, Virginia and New England. The superstorm destroyed many homes, buildings and neighborhoods, leaving millions without power for days and weeks. This storm also effected many data centers and by extension, their customers. Many data centers experienced down time for anywhere from minutes to hours to days, which was very costly. A few data centers managed to avoid downtime all together because they had proper plans in place as well as sufficient backup power supply. Natural disasters can occur anywhere and there is no way for us to predict where the next disaster will occur. By properly preparing in advance with a backup generator for your data center, you can avoid costly downtime and continue to service the clients that rely on you.
Being at ground zero of a natural disaster can be a very scary thing. The last thing you want to be dealing with is trying to get out in the middle of the disaster to try and fix a major problem in your data center, such as downtime, that could have been avoided with proper preparation and planning. When a power outage is experienced in a data center, data loads can be switched over to backup generators until regular power returns and is stable enough to support your data center. When trying to prepare for an unforeseeable event like a natural disaster, data centers need to include generators in their plans and also consider hosting data in multiple locations that are not likely to be effected by the same disaster. As we saw with Superstorm Sandy, sometimes generators fail or there may be difficulty obtaining fuel to keep your generator running for extended lengths of time. The point is, multiple backup plans need to be in place to be fully prepared. According to the Huffington Post, “Experts say failures at data centers can have widespread impact across the Internet, depending on how many customers the data center serves. Still, most companies, particularly smaller startups, don’t back up their virtual infrastructure in multiple places.” Data centers need to have well rounded disaster plans should something unforeseen occur. While it may be costly to have backup generators and to test them regularly to ensure they are functioning properly, the cost of an extended downtime, like many experienced in Superstorm Sandy, will be far more costly.
Having a backup generator for your data center is a best practice that all data centers should employ. Having extended downtime for some companies could effectively put the company at risk of going out of business and if nothing else, could cost a lot of money. We cannot predict when natural disasters will occur or who they will affect. With proper backup power, such as a generator, and a plan in place, data centers can remain functioning properly.
We may have heard the term “mission critical facility” but not be sure exactly what this mean and what all a mission critical facility entails. Just from the title we know that a mission critical facility is very important. A mission critical facility is a building that shelters and supports computer and other ancillary equipment. Failure is not an option in a mission critical facility. It is crucial that you anticipate risk and work hard to prevent any potential problems. According to Ellerbe Becket, the core elements of a mission critical facility are as follows:
• Data center – a centralized repository for the computers and equipment that manage a companys critical applications.
• Networking operation Center (NOC) – a room designed for monitoring and maintaining data center equipment. This is the area where employees supervise server activity, web traffic, network performance, software distribution and related activities.
• The Energy Plant – A section of the facility rpoviding the power and cooling infactstructure to support the data center and Networking Operation Center hardware.
Mission Critical facilities run day and night and thus, they cannot afford downtime. Mission critical facilities must continue to run, even if an element of the facility needs repair. The mission critical facility cannot stop operating just because one element fails or needs to be repaired.
When designing a mission critical facility flexibility is very important. A mission critical facility must be able to adapt to any future changes or needs. Anticipating these potential needs, whether power, capacity or security can be the difference between a well functioning mission critical facility and a major failure of the system.
You must also ensure that your mission critical facility is being constantly maintained so that any potential problems can be anticipated and avoided. Mission critical facilities run on a lot of power, have extreme energy needs and these needs must be able to be met or you could face a system failure. In such a facility, you need to maintain energy efficiency without compromising energy effectiveness and reliability. As more businesses move to running internet companies, more and more businesses are becoming reliant on their data centers and mission critical facilities to ensure that business runs as needed so that revenue is not lost.
With zero disruption to the system being the goal, the design of a mission critical facility is crucial. Potential problems and risks must be anticipated and mitigated. The cost of even a short downtime to a mission critical facility can be hundreds of thousands, even millions of dollars. With such high stakes every precaution needs to be taken so that a mission critical facility can continue to run as it has been designed to.
Downtime is every data center’s worst fear. A data center runs around the clock. With many businesses moving to using the internet to provide customers with constant service, downtime can be a major loss in revenue. But downtime does not only effect the customer, it effects the business, and it’s bottom line, as a whole. Downtime is a loss of both time and money for everyone. In an ideal world for data centers, downtime should never happen. According to Dunn and Bradstreet, “59 perfect of Fortune 500 companies experience a minimum of 1.6 hours of downtime per week.” This results in millions upon millions of money lost per year. Downtime costs can vary significantly in different industries, but ultimately, no industry wants to experience downtime.
When a business is primarily run on the internet, downtime can effectively shut down the operation of the business entirely. To an owner of a web based company, this is simply not an option. If you are a web only retailer or company, you cannot complete sales if the system is unavailable. Lost sales equal many lost or frustrated customers, some of whom may simply go to a competitor’s website to make a purchase. The loss of a sale and of a loyal customer can be devastating to a web based company in the long run.
Even if your business is not entirely based on the internet, downtime can make even the most basic functions of a business impossible to complete. Not only are you paying for the downtime because your business is not functioning, internet sales may not be happening, but also, you are paying your employees who may not even be able to complete the tasks they are paid to do because of downtime.
It may sound extreme that a business could lose millions of dollars on downtime per year but it is a reality and that is why downtime needs to be avoided. The key is to anticipate potential problems, whether it be heating and cooling, power supply or anything else, and mitigate any risks that could be lingering. By investing in a safe and reliable data center you can see a lot of return on your investment. It is important to use a company that can properly design and maintain your data center. By having a properly designed data center you can anticipate heating and cooling needs, make sure there is proper ventilation and power supply and use things like an uninterruptible power supply (UPS) to make sure that even if there is a power problem, you have a backup supply of power to ensure you business continues running as normal. Professional data center design businesses can ensure that you will be able to keep running your business and save you a lot of money down the line by avoiding downtime.
As technology shifts and grows more and more power is used and more attention is being paid to energy efficiency improvement in data centers. Data centers use a lot of power and a lot of costly energy so creating a more efficient data center can save a lot of money without sacrificing power reliability. Energy efficiency is not only important to a business’ bottom line but also the the environment. Data centers use a lot of power and so any way in which you can reduce energy use is a bonus. There are some important steps you can take to design an energy efficient data center.
In a data center, a lot of cooling is needed but even cooling requires the use of additional energy. There are options like using a hot aisle/cold aisle design that will help your systems function better and use less energy to stay cool. You will want variability and flexibility of cooling devices that can be adjsuted to keep everything functioning at an optimal level but will fluctuate to the needs and significantly reduce energy use.
Another important element in data center energy efficiency is that you will want to keep your systems lean and avoid duplication across the board. You can avoid duplication by employing things like cloud-based email. You will also want to turn off unused servers which, in large data centers, can cost thousands per year. More and more companies are moving towards virtualization. It provides the same amount of processesing but on fewer servers, reducing servers reduces energy use.
Also, things like tiered storage can save space by ranking importance and storing older files in a more efficient way. Many companies are offering improved servers that both boost power and have improved energy use.
As with improving energy efficiency in any space from home to office to data center, you will need to elminiate any leaks. For a data center, you will want to eliminate any leaks between your air conditioner and racks. It can be very expensive to ool a room and racks if there are leaks and the air doesn’t make it to the racks.
Improving data center energy efficiency is very important and can save businesses a lot of money in energy costs. According to Gartner, “12% of all data center expenditure is currently consumed in energy-related costs.” That is a major percentage of expenditure for any business and reducing those costs should be of utmost importance. By improving data center energy efficiency you are also doing something positive for the environment which is another benefit that effects everyone.
Designing a data center can be a daunting task to take on. There are many things to consider. You want your data center to have good flow, be energy efficient, and functional for the needs of your business. Some of the primary goals when designing a data center is that it has flexibility and scalability to accommodate fluctuating operational and capacity requirements. Technology is constantly changing and your business needs will also fluctuate so being able to be flexible is a huge benefit. With new technology comes the need for increased power and additional cooling. Also, there can be changes in procedure, safety and security which will also warrant needs for flexibility to accommodate change. If your design is static and cannot adapt to the needs of ever changing technology you will have to spend a lot of money to remedy these problems down the road. If you can anticipate and build in a flexibility to your design you will be able to adapt to any changes needed.
There are a lot of elements to consider when designing a data center and, initially, it can be very overwhelming. Having well qualified professional design your data center is the best option because they can anticipate exactly what you will need based on your specific business. One very important part of data center design is energy efficiency. Data centers use a lot of power and if you do not have energy efficiency you will be spending a lot of unnecessary money. You will also want to ensure that you have proper air management. You can achieve this with optimal positioning of data center equipment with a goal of removing hot air exhaust before it mixes with cool air supply. Some people use the hot aisle/cold aisle design to achieve this. Another very important element of data center design is fire protection and suppression. When working with so much electricity and so much heat you want to ensure that all the money you have invested in a data center will not go up in flames. You will also need a security system. There is a lot of value in the equipment in a data center and you want to make sure you are able to monitor your investment.
The power you need to run a data center can be huge. You will want to asses your overall power requirements and anticipate any shifts in your power needs. A common way to to estimte power and space requirements is to use rack units. You will need to determine whether single phase or three phase power is needed.. And you will most definitely need a good backup generator and Uninteruptible Power Supply (UPS) as well as surge protection to ensure that if there are any power outages or problems, you will not experience downtime, which can be very costly.
By integrating flexibility into your design and anctipating your potential needs you can design an efficient and useful data center. Using a qualified data center design company and taking the initial steps to protect your data center against any potential problems can save you a lot of money and a lot of headaches in the long run.