Sustainability is a key element of information technology within the enterprise, according to a recent article, “Seven Major CEO Concerns CIOs Should Address.” One of the things a CIO can do to control costs and impress the CEO is to look at the “true” cost of running and maintaining the data center.

Electricity and power used to run the corporate data center are typically the number one cost of managing and operating an in-house data center.  Couple power costs with the ineffective design and deployment of data center resources, and you have a perfect storm that can suck money right out of the company’s IT budget.

By metering data center power and gaining insight on where the load is going, you can better understand IT efficiencies and burdens. Also, measuring airflow and server cabinet “heaters,” which cause CRAC units to continually run, will allow you to optimize electricity.

Are you running the AC too cold?  Today’s infrastructure doesn’t have to be kept at meat locker settings like they did in the past. With proper design and deployment you can run (and reduce costs) with less AC than before.

CEOs looking to ensure sustainability are demanding CIOs to get on top of all the latest technologies. As IT executives are working toward implementing new technologies, they must still deliver high-performance computing and network availability.

Implementing sustainable best practices and infrastructure improvements can add to the bottom line. The data center has lots of hidden costs to address – power, cooling, connectivity, disaster recovery. Perhaps this is why more companies are looking to colocation as way to allow more time to focus on improvements that drive business results, while someone else manages the network and ensures uptime guarantees.

Maintaining a disaster recovery worksite is easy. In the case of a disruptive event, moving operations and people to a DR site can be chaos without solid, tested procedures in place. Here’s an inventive idea on how to ensure business continuity procedures are executed flawlessly.

A FIBERTOWN oil and gas customer based in Houston uses an A and B team deployment strategy to ensure disaster recovery operations run smooth and seamless.

This company has an A Team of 15-20 staff members (including IT, operations and HR), who are the “first responders” for any disaster recovery/business continuity deployment.  They head to their DR site – 100 miles northwest Houston – in advance of a storm to prepare for transfer of operations from the corporate office.

The A Team is deployed when a storm is 72 hours off the coast to assemble DR operations. They complete a checklist that includes anything from making sure phone systems are properly routed, organizing docking stations and work areas, and even ensuring there are plenty of coffee and supplies.

While this plan of attack is not uncommon, what’s unique about it is the way this company treats this on-call, A Team staff.

If needed, the company will actually house the families of the staff during a deployment, even finding a kennel for the family pets. The company also has plans in place for a contractor to go to the homes of the families and board the windows and doors. It also has a security contract to monitor the homes while these mission critical staff are away.

DR/BC deployments can be stressful on the business but even more so on the staff as they have to leave family behind to do their jobs.  This company does everything it can to keep the staff worry-free during deployment by ensuring their family (even Rover) is safe.

After the A Team readies the site and a storm is approaching the coast, the B Team will activate. This staff will head to the DR site in Bryan/College Station, set their families up in corporate assigned hotel rooms and activate operations.

For large organizations, this A/B team strategy is an inspiring way to ensure DR operations run without a hitch.

Every business has some kind of insurance these days. We have monetary insurance for theft, lawsuits, natural disasters and employee misconduct. What about business continuity? Why are companies so hesitant to invest in BCP and disaster recovery unless it’s mandated by shareholders or for compliance? If business continuity were looked at as insurance for your business, it would never go unpaid, unplanned or untested.

Think about this in terms of disaster recovery. If your applications and mission critical data are wiped out or even lost for a period of time, is your company out of business? According to the Houston Area Research Center, every dollar spent on disaster preparedness saves seven dollars in recovering disaster-related economic losses.

There is not a cookie cutter business continuity plan for all. Companies should consider best practices while preparing and customizing their own plan.

Get started by building a checklist. Put each of the checklist items in one of three categories – Completed, In Progress or Not Started.

Planning for the impact of an unexpected or catastrophic event on your business

- Identify a coordinator and/or team with defined roles for preparedness and response planning. Potential team members may include: Information Security, Operations, Systems, Police/Security, Physical Plant, Insurance, Legal Affairs, Public Affairs, Personnel Department, Comptroller, Audit Division, Safety Office and/or Emergency Response Team.

- Conduct a business process and services inventory to understand which processes are mission-critical to the survivability of the business.

- Determine acceptable levels of service during the recovery period, and what processes need to be maintained or restored first to keep the business running.

- Identify essential employees and other critical inputs (sub-contractors, services, logistics, etc.) required to maintain business operations by location and function during the event.

- Conduct a technology asset inventory to determine and document the mission-critical technology components, their location, how they’re configured, and who is responsible for management.

- Once key components are identified, determine what measures should be taken to protect and recover them.

- Understand the rules or regulations governing your business operations. If you had a business failure, would you be able to maintain compliance? (Sarbanes Oxley, HIPPA, privacy, etc.).

- Understand customer or business partner performance metrics/service level agreements to assess risk for breach of contract, or to put in place performance remedies for your customers.

- Identify a budget: Quantify the potential costs of downtime or total business failure. Develop a business case to optimally invest in risk mitigation.

Assessing your data and technology needs in the event of a failure in operations

- Determine the status of your existing disaster recovery plan. Do you have one and is it maintained? Have you tested the plan?

- Determine vulnerability of your organization’s technology infrastructure to natural disasters, including hurricanes, floods, fires, earthquakes, etc.

- Set clear recovery time objectives for each of your business/technology areas.

- Determine the need for off-site data storage and backup.

- Develop a technology plan that includes hardware, software, facilities and service vendors.
Secure clear understanding and commitment from vendors on your plan.

- Secure a backup vendor, if necessary, to perform that critical function if your primary vendor is impacted by a business failure

- Perform security risk assessments around specific threats where possible. Examples of data security include: virus protection, intrusion detection, hacker prevention, network events, component failures and systems crashes.

- Assess, if possible and per prior events, how quickly and accurately your business and technology were restored by existing staff. What were the lessons learned so they can be addressed in future planning?

- Determine the effectiveness of your data backup and recovery policies and procedures. Are the procedures fully documented and an appropriate staff member responsible for the maintenance of that documentation?

- Perform a data recovery test. Was the test successful?

- Prepare an incident plan for mitigating a security breach. Audit annually, as security threats can change.

Communicating your plan to employees and vendor partners

- Determine who needs to be contacted with critical information. Build distribution lists and maintain for accuracy.

- Develop a contact plan to reach employees: wireless, home, etc.

- Ensure employees know where to receive information and updates about whether they can return to work, or if they are to report to a different location (Internet, conference bridges, etc.).

- Ensure mission-critical employees know their role in the plan and have access from remote locations (i.e., home broadband, phone, VPN for security).

- Make sure the plan can be executed by alternate employees who are not necessarily the “expert” in cases where those employees cannot be reached.

- Determine the need for a designated recovery site for your people to resume work. Plan for communications, data connectivity, desktops and workspace at that site.

- If you require support from vendor partners, ensure they also have a documented plan that complements your needs. Review periodically to keep the plan current.

Coordinating with external organizations and helping your community

- Collaborate with your local government agency to share your plans and understanding of their capabilities in the event of a business-impacting catastrophe.

- Share your plan with your building management so they have a clear understanding of their role in safely securing the building and your employees.

- Share best practices with other business leaders in your community, chambers of commerce and business associations to improve community response efforts.

For some, outsourcing your data center is like sending a child off to college.

It’s sometimes hard to let go and trust that they are in good hands. You know the experience they’re receiving allows them to run at peak performance, but it can be difficult not seeing them every day. You worry that they may not be safe so far away from home. Added to that…it can sometimes get expensive, until you realize the return on investment is invaluable.

When it comes to your data (or business children) and mission critical IT assets, how do you justify colocation and what’s the ROI?

At first, taking the data center out of your own building and moving it somewhere that could charge you thousands of dollars per month for the next 3-5 years seems complicated and expensive. Upon further investigation though, IT leaders find that reliability, performance and security are worth the consideration.

According to Colocation Industry Trends Report, here’s how IT leaders benefit from colocation:

Reliability & Performance

When running your own data center, you may not be responsible for the electric bill – it just goes to accounts payable and magically disappears.  You may not be aware of the thousands of dollars it takes to keep the data center running at 68 degrees. You may not realize the cost of electricity it takes to power all the new cool blade servers you just bought or that new SAN that you just had to have to store all the valuable data the company needs to run daily operations.

What would you do if you lost power to your SAN and the geologists, bankers, doctors or executives couldn’t access it? How would you answer when the IT manager, CIO or even CEO demands, “Why are we down and how long until systems are back up?” What would downtime cost your business in terms of lost revenue and customer service reputation?

Colocation providers specialize in providing uninterruptable power, high-performance cooling and redundant bandwidth. As a colo customer, you can sleep at night assured your service agreement guarantees uptime and performance so you never have to answer to your executives. How much is that worth?

Security

Third party data center operators watch over customer IT assets as if they were their own. Most providers offer 24×7 security offices with video surveillance, fire and flooding protection, and network monitoring to ensure customer equipment is safely tucked away and running at peak performance. Tier IV data centers guarantee security, reliability and performance. What’s better than a 100% SLA for justifying ROI?

Disaster recovery and business continuity are one of the key elements of colocation. Securing production systems or failover at an offsite location guarantees the continuance of business in the event of disaster.

When it comes to bottom line ROI, some customers see returns in as little as six months. Immediate ROI comes from eliminating the CAPEX of installing and managing hardware in-house. Long term returns come from reducing power and cooling costs, eliminating downtime, ensuring disaster recovery, delivering high-performance bandwidth to users and customers. Most companies should prepare to see a significant improvement in up to 24 months.

Whether you outsource your data center within 10 miles or 100 miles, colocation doesn’t have to be an empty nest experience. It’s the next step in allowing your children (or data) to scale as your business grows with confidence it’s being cared for and operating at optimal performance.

Failing to fully utilize or “load” power circuits to their rated capacity may not result in downtime but could inflate power subscription costs.

Avoid excessive power charges from underutilization through proper power planning and budgeting. This involves loading every circuit to the rated capacity while respecting safety margins.

Many colocation providers deliver power on a subscription basis. In a dual powered data center, the customer pays a flat rate per A-B whip pair. This rate is calculated based on a proportionate cost for the infrastructure required to condition and distribute the power, as well as the electricity consumed at the full 80% load rating. If the circuits are not fully loaded, the customer ends up paying for unused power.

While this may sound like an easy pitfall to avoid, in application, it is often ignored. This is especially true when migrating from a legacy owned facility to a modern colocation facility. Customers simply specify a set number of circuits for each cabinet while neglecting to analyze and evaluate their exact needs.

This can be a costly mistake resulting in leasing more square footage than is absolutely necessary to support the load at the facility’s design power density while paying for subscribed power that is never used.

Another common pitfall is specifying the number of required circuits based on the number of PDUs or power trips in the cabinets. Data center power circuits are an expensive way to handle power distribution and care should be used to order only what is required.

FOR EXAMPLE – When you have gear with low power requirements…

Solve the dilemma of loading each circuit to its rated capacity by using double gang outlets on the ends of the power whips. For example, a 30 amp whip can be outfitted with two 15 amp receptacles in a double gang box. Provided the PDUs plugged in to the outlets are outfitted with 15 amp circuit breakers, this could result in an economical way to safely deliver 15 amps to two different cabinets on a single 30 amp whip pair.

While dual powered facilities provide concurrent maintainability and high availability when coupled with dual powered devices, proper power distribution planning is still required to achieve the ultimate goal of zero downtime and realize the cost benefits of outsourcing.

It’s generally a good design principle to disallow the use of computer devices in a high-availability data center environment. However, some network products or legacy servers may only have single power supplies. This “single-corded” gear will require an automatic transfer of A-B switch to protect the device. Failure to do so during the design phase may lead to power loss and resulting downtime. Here’s the secret to preventing power loss on single-corded gear.

Legacy servers with single power supplies may be performing mission critical functions today, but are destined for upgrades. Until those upgrades are completed, they’re still required to be in service. This is also true of many mid-range network devices such as firewalls, network clocks or NTP servers, and even some edge switches.

These single power supply devices can still be used with reliability by utilizing automatic transfer switches also commonly called Automatic Transfer Power Distribution Units (PDU).

These low-cost devices are typically rack mountable and occupy 1U or rack unit of space. They feature dual input cords and are able to switch from one power circuit to the other in a few micro seconds when power failure is detected on one of the input leads. This transfer time is typically well within the specification of most devices, so the blip is not seen by the load. The power fails, the load transfers and the attached devices continue operating normally.

Once power returns, everything returns to normal. The same design criteria must be observed with the automatic transfer switches as dual power supplied computer devices. Limit the loads to 80% of the breaker rating and split that 80% evenly between the two power circuits.