In order to help, we will try to highlight a few of the points to consider when selecting . . . → Read More: Choosing a Collocation Data Center Facility]]> For many enterprises, placing their computers, storage and networking equipment in a collocation facility makes good business sense. Selecting the right collocation facility that will work for the enterprise can be more complex than many imagine.

In order to help, we will try to highlight a few of the points to consider when selecting the facility that is right for the customer’s individual situation. For this exercise we’ll assume an enterprise is considering a caged space to house between five and 25 cabinets.

Power: Most collocation facilities make the majority of their revenue from power. There are two basic power options available at most collocations. Power is either sold by the circuit or on a metered basis. When purchasing power on a metered basis customers pay for exactly what they use. The rate for metered power may appear expensive since the rates cover a number of collocation overheads along with a fair profit. Although the price per watt is higher, the benefit of metered power is that the invoice reflects actual usage rather than capacity.  This is a good choice for customers when power needs fluctuate or they are running a disaster recovery site which sits idle for long periods of time. Unfortunately unless the enterprise is a large customer most collocation facilities do not offer metered power.

Purchasing a power circuit is based on a flat rate price for the capacity of the circuit regardless of the amount of power drawn on the circuit. In the USA, power circuits are sold on a deregulated basis which means customers will be limited to only 80% of the listed circuit. In other words if a customer purchases a 20 amp circuit, it only has 16 usable amps from that circuit. Two factors impact the real cost when purchasing power on a circuit basis. The first is actual circuit cost.  The second is the customer’s ability to maximize its use of the circuit. For example a customer is paying for a 20 amp circuit, it is limited to a deregulated 16 amps. If the equipment is only drawing 12 amps off the circuit the customer is paying for four amps of unused capacity. Most collocation customers do not fully utilize their power circuits and ultimately pay a higher price for power.

Another factor is a dual power or fail-over circuit. Most computer, network and storage equipment allows for dual power feeds. Purchasing a two power circuit configuration where one will remain unused unless a failure occurs can be an expensive insurance policy. Most collocation facilities will provide backup circuits at half to a quarter of the full price. Secondary power circuit prices are highly negotiable.  When ordering a fail-over circuit, ensure the facility delivers it from a different feed than the primary circuit.

collocation hot aisle

Structural loads: High-density, smaller footprints equal lower costs and allows for easier expansion. Most facilities will market their watts per sqft but few list their floor’s load ratings. If an enterprise deploys high-density racks they need to know both. A full cabinet can easily exceed a ton.  Customers do not want their deployments to be limited by the weight capacity of the facility’s floor.

Network: There are two different models for providing Internet service in collocation facilities. One is carrier neutral and the other is carrier specific. Carrier neutral data center facilities offer customers more choices, price competition, and are generally available in the US and Europe.

In a carrier neutral site, customers will purchase network circuits directly from one or more of the network providers hosted by the collocation facility. The collocation facility will charge a monthly “Cross Connect” fee for maintaining the connection from their network meet-me room to the customer’s cage space. The monthly cross connect charge is negotiable. Facilities will be more likely to discount the cross connect charge if several circuits will be maintained over a number of years.

Carrier specific facilities require customers to purchase network circuits from them directly. This requirement can be expensive and limits cost saving through network techniques such as lowest cost pathing. This type of facility is typically operated by a network carrier such as AT&T, Verizon or other telco.

Infrastructure: Many collocation facilities brand themselves as Tier I, II, III or IV based on the Uptime Institute’s scale. While this rating provides a general impression of the facility, customers should dig deeper. Was the facility purpose built as a data center or is it a converted office space?  Customers should verify a facility’s electrical one lines, UPS and battery strings, HVAC, egress and ingress points for hidden single points of failure. Knowing the facility’s strengths and weak points allows for a better business decision.

Security: Is there a 7×24 guard present? Are there multiple levels of security including man traps, biometric or card access, physical sign in with a guard? Are there cameras and do customers have access to video feeds of their space? Are the security access processes well defined and tightly enforced? How is vendor access handled for customer equipment maintenance or upgrades? Collocation facilities are shared with strangers and the facilities need to protect their customers from malicious and accidental incidents. Customers should deploy lockable cabinets within their cage space to further limit access.  Customers may be required to meet their industry’s specific security standards. Validate these security requirements against the facility’s security procedures to understand what additional measures must be taken to meet the standard.

Remote hands: Collocation facilities are strategically located in cooler climates, close to low cost power and away from potential environmental and terrorist threats. This often puts the facility far from its customers. Savvy customers utilize a lights out strategy for managing their equipment in these remote sites and rely on the facility’s support staff for emergencies as well as regular maintenance. What services are available for smart hands to perform onsite duties? What is the rate for remote hands? Is there a minimum charge? Are the services available 7×24? Staff experience and capabilities will vary widely across facilities.

Cleanliness: Is the overall appearance neat and clean? How do the the spaces of the other customers appear? Are they a mess filled with cardboard and trash? These customer spaces are real fire hazards. Dirt and air contaminants from these spaces will find their way into other customers’ equipment. Customers are relying on the collocation facility managers to police the entire site. Other customers’ space is often a reflection of how the facility’s own infrastructure is maintained.

On-boarding: Once the contract is signed the collocation facility has a number of on-boarding items to complete including running power and network circuits, erecting the cage fence and completing security access. The duration and cost associated with on-boarding can vary widely and is negotiable.

Miscellaneous factors: These factors often can make the difference when selecting between similar collocation facilities.

• Shipments: How is equipment moved into a customer’s cage space? Is there a loading dock available for customer use? Will the collocation facility receive and store shipments or do customers have to be there to receive deliveries? How difficult is it to move equipment from receiving to customer’s cage space (freight elevators, pallet jacks and cart availability, wide hallways…)?
• Run Book: A well run collocation facility is maintained through its run book. The run book defines processes for every situation including routine maintenance as well as emergencies. Is the run book available for review?
• Ceiling height: When customers are running cables in ladder racks, nine foot ceilings will cause a number of challenges. Low ceilings may limit cabinet placement if sprinkler heads become obstructed by cabinet tops or ladder racks.
• Lighting:  Is there good lighting in the space?  Will staff be able to clearly see what they are doing when working on equipment in the cabinets.  Accidental incidents have occurred because staff could not accurately read labels or see port numbers.
• Shared space: Is there a build room set aside for customer use to uncrate and setup equipment prior to moving it into the cage space? Is there a break room for customers to use when on-site? Is mobile phone coverage available in the data center or shared space?
• Disposal: Bringing up a new cage space generates a lot of waste. Does the facility support recycling of cardboard, paper, plastic? How easy is it to dispose of trash?
• On the street: Is the facility capable of supporting planned growth? What is the facility’s reputation? What are current and past customers saying about the service?

While this is not an exhaustive list, it does provide important points that should be considered when searching for a collocation facility.  When selecting a collocation facility, ask the questions: “Is this how I would design a data center?”.  Metagyre assists clients in the selection process to ensure their collocation choice supports the goals of the business and provides the best value for client specific needs.

As services expanded, Client . . . → Read More: New Data Center Services]]> About Client
Client delivers digital entertainment to consumers around the world. Client gave the Internet a voice in 1995 with the first audio streaming solution for the Internet, streaming audio. Continued innovation led to generations of streaming video and its desktop player, which is downloaded millions of times per week.

As services expanded, Client required two data centers with close proximity to a new business partner. Metagyre was engaged to design and manage activities related to the new data centers in Amsterdam and Frankfurt.

This project consisted of building out two new collocation centers, deploying network, server and storage equipment and bringing them on-line. Each data center supported a physical production environment and a virtualized test environment. Client uses a number of technologies to deliver world class services to their customers. This technology includes: EMC storage, Cisco, Checkpoint, Citrix, Dell, VMware, Linux, MySQL, and Oracle.

Design and Build Out

• Design and build out the two new data center facilities including electrical, structured cabling and internet connectivity utilizing collocation facility staff.
• Manage the procurement, shipping and deployment of network, storage and server equipment into the new locations.
• Schedule collocation facility activities for equipment receipt and storage prior to deployment.
• Design rack layouts, server placement, power distribution and port assignments.
• Schedule and manage all professional services and sub-contracts for activities requiring third party activities or equipment purchases. Professional services included the deployment and initial configuration of new storage services.
• Schedule and manage client staff to perform the configuration of new networks, storage assignments, physical server and virtualization services.
• Plan application and data tasks for all applications within the two data centers including the test plans for each.
• Secure stakeholder buy off for each component (owned and shared by multiple stake holders).
• Test core data center infrastructure, redundant power and networking.
• Stage, test and verify networks, servers and storage as operational.

Data Center Build out
Metagyre performed all design and procurement activities on location in Seattle while working closely with client staff to ensure buy off from stakeholders across numerous departments and external business partners. Because of the remote location and Client’s desire to perform the build with their own staff, Metagyre performed several dry runs with the Client team to guarantee all scheduled activities were understood and could be performed within the desired time frames. Additionally, Metagyre managed the equipment procurement and shipping process to reduce delays imposed by customs and other European import regulations.

During the on site deployment, Metagyre managed the teams remotely and provided technical and logistical assistance while reporting progress to stakeholders. Vendor involvement was tightly synchronized by Metagyre to make sure remote staff time was optimized and required vendor support was available.

Following the build out and resiliency testing, the Client application development team deployed the application stack into the test and production environments.

Obstacles
These data centers represented new business ventures for Client with very tight budget constraints. In order to minimize the infrastructure investment and maximize financial return, Metagyre led numerous iterations of design sessions with client engineers. Our understanding of the complete data center stack allowed us to ensure the network, storage, server and infrastructure provided a complete solution within the business and budgetary constraints.

Project Summary
Client’s new data centers were successfully implemented in the originally budgeted dollar amount and time frame. By carefully running activities across diverse teams and vendors, Metagyre was able to complete the project in a heavily matrixed organization. Following the deployment, Client has continued to expand its online services.

Raise the temperature in your data center a few degrees and you can save money.  However, if the temperature in the data center causes the server fans to run too high, . . . → Read More: Save Money With Higher Data Center Temperatures]]> Research presented at the Data Center Energy Efficiency Summit (DCEE) in Sunnyvale, CA. left those in attendance with one major take away. 

Raise the temperature in your data center a few degrees and you can save money.  However, if the temperature in the data center causes the server fans to run too high, they will negate the saving from the chillers and air handlers.

Recommendations form industry groups for heating and air conditioning professionals recommends a temperature range from 77 to 80 degrees.  As the room temperature rises, the  chiller water set point can be raised as well.

The major take away: There are a number of factors that go into reducing energy costs in a data center.  At Metagyre our professionals focus on saving clients money in the daily operations of their data center through improved power utilization,  computing efficiencies, targeted air flow and the appropriate use of proven technologies.

One of the most quoted portions of the standard is . . . → Read More: Collocation Facilities And The Myth Of The Tiered Data Center]]> A number of years ago the ANSI/TIA-942-2005 came out as “the” Telecommunications Infrastructure Standard for Data Centers.  Its purpose was to provide requirements and guidelines when designing a data center and to encourage planning for the data center earlier in the property development process.

One of the most quoted portions of the standard is the “Tiered Reliability”, which was intended as a means for determining specific data center needs required to lessen the time a data center was taken off-line.  The tier breakdown assists designers with a method to quantify certain aspects and to objectively compare one center to another.

Tier I – Basic Components

• single path for power and cooling with no redundant components (N)
• susceptible to disruptions from both planned maintenance activities and unplanned events
• may or may not have back up generator, UPS (uninterrupted power supply) or raised floor
• 99.671% availability – 28.8 hours (statistically) of annual downtime

Tier II – Redundant Maintainable

• single path for power and cooling with redundant components (N+1)
• less susceptible to planned and unplanned disruptions.  Planned infrastructure maintenance may require a processing shutdown
• includes generator, UPS, and raised floors
• 99.749% availability – 22.0 hours (statistically) of annual downtime

Tier III – Concurrently Tolerant

• multiple paths for power and cooling with only one active at any time (active/passive – N+1)
• planned infrastructure maintenance can be performed without disruption.  Unplanned events may cause disruptions
• includes raised floor, generator, UPS with each power and cooling path capable of carrying the full data center load
• 99.982% availability – 1.6 hours (statistically) of annual downtime

Tier IV – Fault Tolerant

• multiple active power and cooling distribution paths including redundant components (2(N+1) S+S) for example each path has UPS in an N+1 configuration
• can sustain planned maintenance activities and at least  one unplanned event without a critical impact on load
• includes raised floors, multiple generators, multiple UPS and other multiple distribution components
• 99.995% availability – 0.4 hours (statistically) of annual downtime

The standards went a long way to elevate the need for specific elements in the data center and the best practices for their usage.  Unfortunately, once the marketing spin took off, numerous collocation vendors began touting their facilities as a Tier III or Tier IV without regard for the full spectrum of component requirements or the processes supporting those components.  In addition the design standards cover other elements including:

• cable and fiber designs
• network distribution
• rack and cabinets layout
• space utilization and pathways
• air flow design

Many commercial collocation facilities claiming higher fault tolerance, achieve this mark by maintaining a lower power density requirement on their tenants driving up customer costs and reducing cabinet layout efficiencies. Even when the power plant supports high electrical draws, floor load ratings may limit the cabinet’s capacity.  Other times, best practices must be completely rethought, for example the best hot and cold aisle layout will fail to achieve design goals if the collocation facility locates the client in a cage space which does not support air flow management. Of course occasionally collocation customers themselves lower the reliability by failing to incorporate proper redundancy practices within their private suite or cage space.

Each business has unique up-time, cost, networking and processing density goals in mind when they establish a data center.  In order to reach those goals, a balance between requirements and design standards must be struck.  While the tiered data center may be an over used marketing myth, the ANSI/TIA-942-2005 provides a solid base template.  This template, along with skilled engineering and management practices can go a long way to facilitate a data center build out that delivers on technical design goals at an agreed upon business price point.

Partnering is not a new sales term for the old cost-focused . . . → Read More: Data Center Management: Partnering For Success]]> At its best, “partnering” with a vendor is a powerful tool that can drive successful business outcomes for data center managers. Done right, partnering can deliver significant value for the money to both parties over the length of the relationship.

Partnering is not a new sales term for the old cost-focused customer verses supplier management battle. Partnering involves working collaboratively with vendors. Its intent is to distribute business and technology risks to the partner who can most successfully manage those risks. Success involves both partners, the customer and the vendor, realizing benefits once shared business objectives are met. Those business objectives are defined in terms of desired outcomes rather than specific deliverables or traditional service level agreements (SLA). In a successful partnership, the relationship itself is an asset with value to both parties.

To realize success, data center managers and vendors have to embrace a greater degree of openness, communication, and trust. Partnering involves the customer and the vendor working cooperatively to identify optimum solutions. Partnering can succeed only where there is a mature approach to relationship management and where effort and commitment, from both parties, are invested to make the relationship work. Partnering should never be regarded as an easy option.

Like most changes in an enterprise, successful partnering starts with the senior management team placing the full weight of their positions behind the desired relationship and agreeing on clear objectives. This commitment includes assigning the best people to create and maintain the relationship over a multi-year engagement.  Additionally, significant resource requirement commitment must be made to the partnering arrangement. This commitment includes staff and investment capital that may not pay off for several years. In a partnering relationship, the vendor must be willing to expose profit areas in open dialogs with the customer.  Customers must be expose current and future budgets as well as the process for acquiring corporate funds.

While partnering is a powerful tool for data center managers, it is not suitable for every situation. It is most appropriate for activities where there is a need for:

• continuous innovation to address changing business needs over an uncertain future
• staff flexibility with specialized skills and knowledge
• outsourcing business processes or technology services in order to free internal staff to focus on core areas
• development of a new service or system

Conversely, partnering is much less successful when there is a requirement for:

• short-term, stop-gap fixes
• little flexibility in the approach
• tight control by the customer
• activities with limited ability to improve processes or provide innovation
• the vendor to take on an unmanageable risk

In the practical sense, managing a partnering agreement comes down to the customer-vendor relationship. Both parties need to consider how they will build trust, share risk, agree on accountability, and share success. There has to be continual conversations focused on reasonableness. How will success be demonstrated and compromise achieved without the burden of an overly taxing bureaucracy?

Establishing a partnering approach requires both parties to openly discuss what they want and need from the contract. In the early stages, dialogs between the customer and vendor about goals are unlikely to align well. Throughout the process, each party will face compromise and need to stress their willingness to remain at the table in a collaborative fashion.  Without setting a negative tone, an exit strategy should be developed from the outset. Both parties need to understand what is at stake and how the contract will end successfully as well as prematurely.

Establishing a meaningful relationship requires cultural and behavioral changes. Trust is earned through positive demonstrations over time. There will be a steep learning curve for the staff of both parties. Climbing this curve translates into having the right people in place. These people need to be able to make difficult decisions and have the authority to remove staff and obstacles which are hindering the relationship. Where customer and vendor staff are working together, peer relationships need to be established. Throughout the chain of command the partnering relationship must be reinforced. Initially, if the people will not change, change the people.

One consideration to be addressed throughout the chain of command is: how open should each side be with the other? Questions to consider include:

• Shared incentives – are increased savings or revenue for the customer translated in to a profit sharing opportunity for the vendor?
• Project control board – will there be active involvement from senior management from both parties in controlling scope, managing progresses toward goals and resolving issues? Are both parties represented as equals?
• Shared problem, issue and risk logs – will there be a single set of shared records for activities tracked and addressed by both parties in order to avoid duplicate efforts and focus shared resources on a single set of activities?
• Staff reporting – how will roles and responsibilities for managing staff be defined, regardless of affiliation? Who will direct resources and assign tasks?
• Open project accounting books – are both parties invited to review the projects accounting books? Will everyone see budgetary constraints and cost considerations of each business decision in the project life cycle? Is everyone able to see value for their money without focusing on the lowest cost?
• 360 staff & management feedback – will both parties contribute to the feedback loop? How will staff and managers discover what is missing and close the gaps?

The main advantage of this openness is the ability to raise issues much earlier and address them with a shared sense of ownership and responsibility. Resolving problems earlier, when they can be more easily solved, avoids unnecessary costs, time and blame.

While all partnering arrangements start with a contract there is usually a “partnering charter” as well. The charter states desired outcomes and goals that complement the terms and conditions of the contract. Creating a partnering charter usually identifies the principles, attitudes and ideals of the partnership. The charter will also set the level of openness and the code of behavior. It should be consistent with the procurement practices of both parties.

Once the contract and partnering charter are in place the relationship work begins. A series of team building activities need to take place with active involvement from everyone from both parties, including senior management. These exercises will express to everyone the incentives for both parties as well as the expected behaviors which will deliver these incentives.

If all this sounds simple, it can be. There are also common pitfalls. One of the most common mistakes is to have individuals in crucial roles that are not suited to partnering. Most often interpersonal skills are more critical than technical understanding. Organizational maturity is another area often needing to be strengthened. If either organization is missing the cultural maturity to make partnering work, the changes required by may be out of reach regardless of how clear and achievable the benefits. When objectives are unclear, no amount of effort will make the relationship successful. One of the more difficult components of partnering is the articulation of adequate performance measurements. Finding relevant benchmarking measures that make meaningful comparisons between current and desired outcomes can be hard to achieve.

How do you know if your enterprise is ready for a partnering relationship with a vendor? The first step is to take an honest look at your enterprise and ask:

• how has the enterprise worked with vendors to date? Are vendor relationships aimed at driving to the lowest price? Are vendors viewed as adding value? Is there a shared responsibility for success? How is the enterprise viewed in the vendor community?
• is the right leadership in place with the skills and capabilities to work with the vendors collaboratively? Is the management style of the enterprise matrixed with shared responsibilities or siloed and highly controlling?
• has the enterprise clearly defined desired successful outcomes? Are measurements used in current projects to course correct? Are project targets and milestones currently published to all levels of the enterprise?
• is change easily instilled in the enterprise? Is there a successful track record of change within the enterprise?
• are mistakes viewed as opportunities to learn? Are defects found in testing viewed positively or considered a problem? Is quality about capturing metrics or improving outcomes? Is constructive criticism used to improve the process?

Assuming the relationship can be created successfully, will stakeholders ‘sign up’ to it and add momentum to its development? What kind of vendor could manage the risks envisioned to be allocated to them? Realistically, would a vendor be willing to take on the risks? If so, can the enterprise give the vendor sufficient control so that they can manage those risks?

Selecting a vendor who will become a partner means establishing goals that are compatible and achievable to mutual benefit of both parties. The selection of the right partner is not at the expense of an open competitive selection. The final choice of a partner is based on mutual benefit, cultural match, risk management and value over the life of the agreement.

Partnering is not right for every customer, vendor or engagement. At the very least, considering a partnering relationship allows data center managers, project managers and vendors to seriously evaluate their respective strengths and weaknesses as well as the risks and rewards at stake. Under the appropriate circumstances, partnering can deliver significant value over the length of the relationship for both parties.

You only relocate a data center once (hopefully), but you manage it day after day. A big part of making your data center more manageable, dressing in all the cables and fiber.  This means keeping everything covered, bundled and out of sight:

a) Cabinets . . . → Read More: Data Center Design: Keep It Covered, Bundled And Out Of Sight]]>

You only relocate a data center once (hopefully), but you manage it day after day. A big part of making your data center more manageable, dressing in all the cables and fiber.  This means keeping everything covered, bundled and out of sight:

a) Cabinets have doors and sidewalls.
b) Copper cables are patched within the cabinets using the provided cable management.
c) Cables follow standard paths and do not cut across cabinets.
d) Fiber and copper cables are kept in separate trays.
e) Blanking panels will be used.

Dressing in the miles of cable and fiber required to connect all the data center equipment is an art form. With a little practice you will learn how to comb through the cables to ensure they are aligned and bundled so they stay out of the way and allow access to the hardware they connect.

A large part of a well dressed facility is having a solidly designed standardized cabinet layout along with a good structure cable design and power plan. Obviously each category of equipment will require its own layout and each should be considered and planned before any equipment or cable is ever brought into the data center.

All your designs will be a compromise between capacity, initial cost, and management. Since a majority of a data center’s TCO (Total Cost of Ownership) results from the ongoing management and maintenance, spend the time up front and perform the necessary work to keep everything covered, bundled and out of site. You will see the savings over the life of your data center.