Saturday, September 27, 2014

Data center Facility management, saatnya berubah

 February 4, 2014 2 Comments »
Asset management can have a significant impact on the operational performance and profitability of an asset intensive organization so it has become a hot topic in recent years. It is no longer about fixing assets when then break but rather about employing cost effective asset management strategies that maximize asset availability and reliability by minimizing the probability of system failures. Effectively executed asset management can increase the economic life of capital equipment, increase system reliability and reduce maintenance related costs.

Reactive Maintenance

Keeping equipment working is paramount to any data center organization so moving to an asset management philosophy can increase the life of capital assets such as HVAC systems, UPS, generators and buildings. This can be difficult to achieve if the IT way of doing maintenance has become entrenched in the culture of the data center organization. When IT hardware fails, technicians react to the situation and swap out the hardware. It is called “reactive maintenance”. Many datacenter organizations have adopted this reactive way of thinking when it comes to supporting data center infrastructure. If it breaks, let’s fix it! Relying exclusively on reactive maintenance is like flying blind, however, reactive maintenance is still the predominant way of working in North America. Approximately 55% of maintenance activities in the average facility are still reactive, and in many datacenters this is significantly higher. Maintenance costs may make up 2% of data center annual budgets but there are a number of other indirect costs that must be considered.

Indirect Costs of Reactive Maintenance

Indirect costs include data loss or corruption, damaged reputations, higher energy costs, equipment damage, collateral damage, inefficient use of resources, safety issues, legal and regulatory repercussions and difficult to control maintenance budgets. For example, clogged filters can increase the load on blower systems, shorten the life of the motor and increase energy consumption by 15%. The main reason for catastrophic data center facility failure is poor electrical maintenance. Sporadic ongoing repair leads to asset deterioration, a shorter asset lifetime and increased long-term capital cost. The first step on the road to asset excellence is to change the mindset from a reactive to proactive way of thinking.

Planned Preventive Maintenance

Planned Preventive Maintenance (PM) is taking precautionary and proactive steps against unscheduled equipment downtime and other avoidable failures. The purpose of the PM is to institute scheduled maintenance and inspections so that defects can be spotted before they evolve into something more severe. Regular preventive maintenance is critical to maintaining the reliability of your data center infrastructure. It will keep your CRAC, HVAC and backup generators in optimal working condition. Sample Preventive Maintenance Tasks – Replace HVAC filters – Check CRAC temperatures – Verify oil levels in compressors – Lubricate moving parts such as bearings – Verify backup generator is operating ok – Inspect fire safety systems Planned maintenance reduces the chances of emergency repairs and downtime; it leads to an increase in system reliability; delivers energy and efficiency improvements resulting in a net reduction in maintenance costs. It is important to complete these PM’s and to complete them on time. This is difficult to achieve if you are relying on paper, spreadsheets or wiki software. How do you determine what planned maintenance is due, who did what, what needs to be done etc without an automated system?

Track Your Data Center Maintenance in a CMMS

Creating a data center preventive maintenance plan can be accomplished quickly with a CMMS. Human error is often cited as one of the leading factors in data center downtime so moving to a CMMS system to track your data center maintenance is the first step on the road to continuous improvement. Standardized procedures and MOP’s can be tracked digitally in the CMMS and triggered when due using the scheduled maintenance tool. Work order information can also be used to identify chronic equipment problems and unacceptable levels of downtime, so solutions such as regular inspections or preventive maintenance can be put in place to proactively reduce the level of downtime going forward. Transitioning to a CMMS can be quick and painless with modern cloud based CMMS. Most contemporary CMMS applications have import tools and copy functions to get your data setup and configured quickly. In 2013, we assisted 365 Main in moving toward a “predictive whole facility view” of data center maintenance. Today, they are saving up to 40 man-hours per site per month on just managing  basic maintenance processes with a CMMS.” Read how there did it here.
About the Author:
Jeff O’Brien is an product specialist and blogger at Maintenance Assistant Inc, a leading provider of innovative web-based CMMSCMMS software is the ideal way to manage your maintenance on facilities and equipment. You can view their website at or reach Jeff on Linkedin.

7 tips mengelola perawatan data center Anda

Jeff O’Brien is an industry specialist and blogger at Maintenance Assistant Inc., a provider of innovative web-based CMMS, which is a tool to manage facilities and infrastructure equipment at data centers. 
Jeff-O'Brien-tnJEFF O’BRIEN
Maintenance Assistant
Can you afford to have one of your critical power distribution assets fail because you missed your scheduled preventative maintenance? According to a recent study by Ponemon Institute, one minute of data center downtime now costs $7,900 on average. With an average reported incident length of 90 minutes, we can calculate that the average incident now costs $700,000. This large cost is related to the fact that modern data centers are supporting critical websites and cloud software applications. 
Preventative maintenance ensures maximum reliability by taking precautionary and proactive steps to reduce unscheduled equipment downtime and other avoidable failures. The purpose of preventive maintenance is to institute scheduled inspections so that defects can be spotted before they evolve into something more severe. Those data centers that don’t carry out planned and preventative maintenance have an increased risk of asset failure. Here are seven tips for first class preventive maintenance at your data center:

Safety First

Data centers contain numerous hazards that can impact the life and health of technicians. Data center technicians need to be aware of potential safety hazards when performing preventive maintenance activities. Use lockout tag out where necessary. Ensure data center technicians are familiar with health and safety processes when performing preventive maintenance tasks by documenting them in every PM procedure and by providing regular safety training.

Schedule Regular Maintenance and Inspections

Performing preventive maintenance on UPS and batteries greatly reduces the chance of failure during power outages. In a recent study by Emerson Network Power, the Mean Time Between Failures (MTBF) for UPS units that received two PM service events per year is 23 times higher than a machine with no PM service events per year. The same is true for other critical systems such as HVAC and generators. Regular preventive maintenance can reduce the chance of failure, reduce the amount of energy consumed and extend equipment lifetime. The manufacturers recommended preventive maintenance is a goof place to start. These can be revised over time.

Use Standardized Checklists

Standardized checklists ensure technicians know what to do during the preventive maintenance, while ensuring the same standard checks are being performed every time. Again, the manufacturers recommended maintenance is a good point to start and this can be refined over time. In data centers, these are called Standard Operating Procedures (SOP), Methods of Procedure (MOP) and Emergency Operating Procedures (EOP). Certain jobs may require additional training and safety procedures so SOPs, EOPs and MOPs should outline what is required.

Enforce PM Compliance

Given the potential costs of data center downtime, it is important to complete PM’s and to complete them on time. The easiest way to do this is to measure and enforce PM compliance. Your preventive maintenance compliance (PMC) score is the percentage of scheduled PM work orders that get done on time. Who decides that time? The 10% rule of maintenance is a good starting point. The rule states that a preventive maintenance action should be completed within 10% of the scheduled maintenance interval. For example, a quarterly PM every 90 days, should be completed within 9 days of the due date or it is out of compliance. The 10% rule can help keep your PM intervals constant, reducing the time variable variation, thus improving reliability.

Keep Detailed PM and Work Order Records

If things go wrong, insufficient documentation can lead more heartache than a data center manager needs. Well-documented PM reports ensure the data is readily available whenever the auditors come to inspect. Also Historical work order information can be used to identify chronic equipment problems and unacceptable levels of downtime so solutions such as regular inspections or preventive maintenance can be put in place to proactively reduce the level of downtime going in the future.


As Facebook CEO Mark Zuckerberg said, “you cannot improve what you cannot measure”. By measuring your performance through KPI’s such as PM compliance, availability and reliability; you can optimize your preventive maintenance to maximize its effectiveness while minimizing costs.

Get a CMMS

A CMMS is the best way to help track, measure and improve your preventive maintenance and meet 6 tips and tricks mentioned above. Computerized Maintenance Management Systems (CMMS), sometimes called Preventive Maintenance Software, enables the facility manager, subordinates and customers to track the status of maintenance work on their assets and the associated costs of that work in one comprehensive system. CMMS software can help a data center drive down the cost of maintenance, increase asset life, improve reliability and productivity and reduce equipment downtime. It ensures that preventive maintenance is performed regularly according to established protocols. It will also give the maintenance technician quick access to equipment information such as procedures, work order history and maintenance data, and metric indicators.
Regular, scheduled maintenance can easily pay for itself by preventing unplanned downtime events thanks to battery or capacitor failure, clogged air filters, welded relays and even outdated firmware. You can digitize and streamline your preventive maintenance operations by consolidating personnel, documents, assets, data, work logs and inventory in one location using a CMMS.

Saturday, September 13, 2014

5 hal tentang data center yang harusnya diketahui Admin

5 Facts about Datacenters Every Administrator Should Know

Being knowledgeable about datacenter basics will help you solidify your technical career. Datacenters large and small operate on a certain set of principles that can be easily adhered to in order to make the most of your time. Any downtime is unacceptable in many business leaders’ eyes therefore building out or leasing reliable datacenter infrastructure is the key to being successful in your technical endeavors.
Downtime Costs $5,000 per minute
Sometimes more and sometimes less depending upon the industry but most businesses that have the need for a datacenter need the services for the exact reason of having to be up and running at all times due to business demands. ThePonemon Institute is credited with coming up with this number representing the cost of datacenter downtime.
Older Servers May Use Less Power
Although data centers themselves are getting greener, the newer datacenters utilize more powerful hardware which may outperform older hardware significantly while taking up more power resources. This may come as a surprise to datacenter veterans who are touching newer datacenter hardware for the first time.
Having HVAC Experience Helps
Most technical related career paths do not take the time of teaching you the basic principles of HVAC. Knowing the basics of HVAC will help you as you build out onsite private datacenters. Keeping your equipment properly cooled is essential to running an efficient datacenter and many datacenter administrators have to learn HVAC terms and basics on the fly when negotiating contracts with vendors. Knowing how to troubleshoot HVAC systems could save you a hefty contractor bill should you encounter a minor issue that require an HVAC tech to come on site to fix.
Datacenters as a Service: Welcome to the new age
Facilities as a Service as they are being dubbed are modular datacenters that can be quickly connected to electricity and cooling equipment while having the bare metal servers inside the datacenter ready to be spun up and staged with whatever cloud or virtualization suite you see fit. Datacenter administrators can order prefabbed modular datacenters that fit the exact specifications of the client.
Computing Loads are Unpredictable
That’s what your professor never told you in your computer science 101 class. If humans had perfect control over computational needs, many of the challenges we face as a human race could be solved. That just isn’t feasibly possible in every scenario. The computing loads that your end users place on the system is totally unpredictable due to the amount of moving parts that are present within most organizations. The best tip is to reduce the amount of moving parts whenever possible in order to reduce the effects of spiking computing loads.

Sunday, September 7, 2014

Google data center gunakan Recycling Equipment

As part of our commitment to keeping our users' data safe, we destroy failed hard drives on site before shredding and recycling them.

Extending our equipment lifecycle

From the moment we decide to purchase a piece of equipment to the moment we retire it, we reduce, reuse, and recycle as much as we can.

We reduce by sourcing locally

Whenever possible, we use local vendors for heavier components like our server racks. Even if material is more expensive locally, we can recoup the extra cost by reducing shipping charges from farther locations where the material may be cheaper. By limiting the shipping distance, we reduce the environmental impact of transportation.

We reuse existing machines

Before we buy new equipment and materials, we look for ways to reuse what we already have. As we upgrade to newer, higher-speed servers, we repurpose older machines either by moving them to services that don’t require as much processing power, or by removing and reusing the components that are still in good working condition.
Since 2007, we've remanufactured and repurposed enough outdated servers to avoid buying over 300,000 new replacement machines.

We recycle as much as we can

At each location, we maximize the recycling of all our data center material. When we can’t repurpose our equipment, we completely erase any components that stored data, and then resell them into the market, giving this equipment a second life. We break down any components that can’t be sold into raw materials like copper, steel, aluminum, and plastics for responsible recycling.

Google data center gunakan Water Cooling

Colorful pipes carry water in and out of the data center. The blue pipes supply cold water and the red pipes return the warm water back to be cooled.

Cooling with water—not chillers

The electricity that powers a data center ultimately turns into heat. Most data centers use chillers or air conditioning units to cool things down, requiring 30-70% overhead in energy usage. At Google data centers, we often use water as an energy-efficient way to cool instead.

We trap hot air and cool our equipment with water.

We've designed custom cooling systems for our server racks that we've named “Hot Huts” because they serve as temporary homes for the hot air that leaves our servers—sealing it away from the rest of the data center floor. Fans on top of each Hot Hut unit pull hot air from behind the servers through water-cooled coils. The chilled air leaving the Hot Hut returns to the ambient air in the data center, where our servers can draw the chilled air in, cooling them down and completing the cycle.

We take advantage of evaporative cooling.

Evaporation is a powerful tool. In our bodies, it helps us maintain our temperature even when outside temperatures are warmer than we are. It also works similarly in our cooling towers. As hot water from the data center flows down the towers through a material that speeds evaporation, some of the water turns to vapor. A fan lifts this vapor, removing the excess heat in the process, and the tower sends the cooled water back into the data center.

We use the natural cooling power of sea water.

Evaporating water isn't the only way to free cool. Our facility in Hamina, Finland uses sea water to cool without chillers. We chose this location for its cold climate and its location on the Gulf of Finland. The cooling system we designed pumps cold water from the sea to the facility, transfers heat from our operations to the sea water through a heat exchanger, and then cools this water before returning it to the gulf. Because this approach provides all of our needed cooling year round, we haven't had to install any mechanical chillers at all.

We save and recycle water.

To better conserve water, we power two of our data centers with 100% recycled water, and capture rainwater for cooling a third. The idea is simple: instead of using potable (or drinking) water for cooling, we use non-drinkable sources of water and clean it just enough so we can use it for cooling.
We use recycled water from various sources. Our Douglas County facility treats city waste water, while our Belgium facility pulls water from an industrial canal. There, we use a large tank filled with fine sand to filter out small particles, leaving the water completely clear (although not suitable for drinking). While it's not always possible or cost-effective to use recycled water, we're optimistic that we can find sustainable solutions for the majority of our water use.

Google data center gunakan Plastic Curtain untuk jaga suhu

Plastic curtains in the network room prevent the hot air behind the server racks from mixing with the colder air in front of the server racks.

Controlling the temperature of our equipment

To help our equipment function optimally while continuing to save energy, we manage the temperature and airflow in our data centers and machines in simple, cost-effective ways.

We raise the thermostat to 80°F.

One of the simplest ways to save energy in a data center is to raise the temperature. It’s a myth that data centers need to be kept chilly. According to expert recommendations and most IT equipment manufacturers' specifications, data center operators can safely raise their cold aisle to 80°F or higher. By doing so, we significantly reduce facility energy use.

We plan by using thermal modeling.

We use thermal modeling to locate “hot spots” and better understand airflow in the data center. In the design phase, we physically arrange our equipment to even out temperatures in the facility. Even after that, we can move certain equipment like computer room air conditioners (CRACs) to reduce hot spots and even out the ambient temperature—ultimately reducing the amount of time the CRAC must run.

We find inexpensive ways to manage airflow.

To cut cooling costs and save energy, we prevent the “hot aisle” air behind the server racks from mixing with the “cold aisle” in front of the server racks. In our large data centers, we use appropriate ducting and permanent enclosures. In addition, we take simple measures well-suited for smaller “closet” style data centers. For instance, we:
  • Use blanking panels (or flat sheets of metal) to close off empty rack slots and prevent hot aisle air from seeping into to the cold aisle.
  • Hang plastic curtains (like those used in commercial refrigerators) to seal off the cold aisle.
  • Enclose areas with components that run hotter (such as power supply units or PSUs) with plastic curtains.
These efforts help to reduce the total amount of energy used for cooling. At the same time, they ensure that the cooler air we send into the cold aisles is truly cool enough to do its job.

Google data center gunakan server custom.

Blue LEDs on this row of servers tell us everything is running smoothly. We use LEDs because they are energy efficient, long lasting, and bright.

Building custom, highly-efficient servers

Google's servers are high-performance computers that run all the time. They're the core of our data centers, and we've designed them to use as little energy as possible. We do this by minimizing power loss and by removing unnecessary parts. We also ensure our servers use little energy when they're waiting for a task, rather than hogging power when there’s less computing work to be done.

We optimize the power path.

A typical server wastes up to a third of the energy it uses before any of that energy reaches the parts that do the actual computing. Servers lose the most energy at the power supply, which converts the AC voltage coming from a standard outlet to a set of low DC voltages. They then lose more at the voltage regulator, which further converts the power supply's output to the voltages required by microchips. Designed with low efficiency standards in order to save on initial cost, traditional servers end up costing much more in electricity in the long run.
Google servers are different. Since 2001, we've been designing them to be more efficient. For many years now, we've used very efficient power supplies. Similarly, we use highly efficient voltage regulator modules to ensure that most of the power goes to the components that do the actual computing work. We’ve cut out two of the AC/DC conversion stages by putting back-up batteries directly on the server racks. We estimate an annual savings of over 500 kWh per server—or 25%—over a typical system.

We customize our servers to stick with the essentials.

When it comes to hardware, we only use what's necessary for our applications to run. We remove unnecessary components like peripheral connectors and video cards. We also optimize our servers and racks to use minimal fan power. The fans spin just fast enough to keep our machines cool enough to run.
We encourage all of our suppliers to produce parts that operate efficiently whether they’re idle, operating at full capacity, or at lower usage levels—an idea we introduced as "energy proportionality" in 2007. Our study shows that more energy-proportional systems can cut total energy use in half at large data centers.

Mengukur efisiensi data center Google.

Measuring and improving our energy use

We're focused on reducing our energy use while serving the explosive growth of the Internet. Most data centersuse almost as much non-computing or “overhead” energy (like cooling and power conversion) as they do to power their servers. At Google we’ve reduced this overhead to only 12%. That way, most of the energy we use powers the machines directly serving Google searches and products. We take detailed measurements to continually push toward doing more with less—serving more users while wasting less energy.

We take the most comprehensive approach to measuring PUE

Our calculations include the performance of our entire fleet of data centers around the world—not just our newest and best facilities. We also continuously measure throughout the year—not just during cooler seasons.
Additionally, we include all sources of overhead in our efficiency metric. We could report much lower numbers if we took the loosest interpretation of the Green Grid's PUE measurement standards. In fact, our best site could boast a PUE of less than 1.06 if we used an interpretation commonly used in the industry. However, we're sticking to a higher standard because we believe it's better to measure and optimize everything on our site, not just part of it. Therefore, we report a comprehensive trailing twelve-month (TTM) PUE of 1.12 across all our data centers, in all seasons, including all sources of overhead.
Data center power distribution schematicFigure 1: Google data center PUE measurement boundaries. The average PUE for all Google data centers is 1.12, although we could boast a PUE as low as 1.06 when using narrower boundaries.

Google data center PUE performance

Our fleet-wide PUE has dropped significantly since we first started reporting our numbers in 2008. As of Q2 2014, the TTM energy-weighted average PUE for all Google data centers is 1.12, making our data centers among the most efficient in the world.
Data center PUE
Figure 2: PUE data for all large-scale Google data centers

Q2 2014 performance

Fleet-wide quarterly PUE:1.11
Fleet-wide trailing twelve-month (TTM) PUE:1.12
Individual facility minimum quarterly PUE:1.08, Data Center N
Individual facility minimum TTM PUE*:1.10, Data Centers J, L, N
Individual facility maximum quarterly PUE:1.14, Data Centers B, C, E, F
Individual facility maximum TTM PUE*:1.14, Data Centers B, F
* We report Individual Facility TTM PUE only for facilities with at least twelve months of operation.
For Q2, our fleet-wide quarterly PUE was 1.11, and our fleet-wide TTM PUE was 1.12. These are the same as the previous quarter's results.
In Q2, we had five sites reporting quarterly PUEs of 1.10 or less. Three sites had TTM PUEs of 1.10 or less in the same period.

Measurement FAQs

What's the average PUE of other data centers?

According to the Uptime Institute's 2014 Data Center Survey, the global average of respondents' largest data centers is around 1.7.

Why does our data vary?

Our facilities have different power and cooling infrastructures, and are located in different climates. Seasonal weather patterns also impact PUE values, which is why they tend to be lower during cooler quarters. We’ve managed to maintain a low PUE average across our entire fleet of data center sites around the world—even during hot, humid Atlanta summers.

How do we get our PUE data?

We use multiple on-line power meters in our data centers to measure power consumption over time. We track the energy used by our cooling infrastructure and IT equipment on separate meters, giving us very accurate PUE calculations. We account for all of our power-consuming elements in our PUE by using dozens or even hundreds of power meters in our facilities.

What do we include in our calculations?

When measuring our IT equipment power, we include only the servers, storage, and networking equipment. We consider everything else overhead power. For example, we include electrical losses from a server's power cord as overhead, not as IT power. Similarly, we measure total utility power at the utility side of the substation, and therefore include substation transformer losses in our PUE.
Data center power distribution schematicFigure 3: Google includes servers, storage, and networking equipment as IT equipment power. We consider everything else overhead power.

Equation for PUE for our data centers

PUE equation
ESIS Energy consumption for supporting infrastructure power substations feeding the cooling plant, lighting, office space, and some network equipment
EITS Energy consumption for IT power substations feeding servers, network, storage, and computer room air conditioners (CRACs)
ETX Medium and high voltage transformer losses
EHV High voltage cable losses
ELV Low voltage cable losses
EF Energy consumption from on-site fuels including natural gas & fuel oils
ECRAC CRAC energy consumption
EUPS Energy loss at uninterruptible power supplies (UPSes) which feed servers, network, and storage equipment
ENet1 Network room energy fed from type 1 unit substitution

We’re driving the industry forward

We’ve made continual improvements since we first disclosed our efficiency data in 2008. As we've shared the lessons we’ve learned, we're happy to see a growing number of highly efficient data centers. But this is just the beginning. Through our best practices and case studies, we continue to share what we’ve learned to help data centers of any size run more efficiently.

5 cara Google mencapai efisiensi data center nya.

  1. Measure PUE

    You can't manage what you don’t measure, so be sure to track your data center's energy use. The industry uses a ratio called Power Usage Effectiveness (PUE) to measure and help reduce the energy used for non-computing functions like cooling and power distribution. To effectively use PUE, it's important to measure often. We sample at least once per second. It’s even more important to capture energy data over the entire year, since seasonal weather variations affect PUE. Learn more.
  2. Manage airflow

    Good air flow management is crucial to efficient data center operation. Minimize hot and cold air mixing by using well-designed containment. Then, eliminate hot spots and be sure to use blanking plates (or flat sheets of metal) for any empty slots in your rack. We've found that a little analysis can have big payoffs. For example, thermal modeling using computational fluid dynamics (CFD) can help you quickly characterize and optimize air flow for your facility without having to reorganize your computing room. Learn more.
  3. Adjust the thermostat

    The need to keep data centers at 70°F is a myth. Virtually all equipment manufacturers allow you to run your cold aisle at 80°F or higher. If your facility uses an economizer (which we highly recommend), run elevated cold aisle temperatures to enable more days of "free cooling" and higher energy savings. Learn more.
  4. Use free cooling

    Chillers typically use the most energy in a data center's cooling infrastructure, so you'll find the largest opportunity for savings by minimizing their use. Take advantage of "free cooling" to remove heat from your facility without using a chiller. This can include using low temperature ambient air, evaporating water, or a large thermal reservoir. While there's more than one way to free cool, water and air-side economizers are proven and readily available. Learn more.
  5. Optimize power distribution

    You can minimize power distribution losses by eliminating as many power conversion steps as possible. For the conversion steps you must have, be sure to specify efficient equipment transformers and power distribution units (PDUs). One of the largest losses in data center power distribution is from the uninterruptible power supply (UPS), so it's important to select a high-efficiency model. Lastly, keep your high voltages as close to the power supply as possible to reduce line losses. Learn more.