April 12-15, 2010 · Rosen Shingle Creek, Orlando, FL

“Best Practices in Storage” Awards Program – Fall 2009

Sponsored by:

Congratulations to our Honorees and Finalists!

The top honoree in each category were announced on the main stage during an Awards Ceremony at SNW, October 13th. All finalists will be recognized on the event website, in an event press release, and may be featured in a special supplement of Computerworld magazine.

[ Download Awards Ceremony Photos 1-13 Zip ]
[ Download Awards Ceremony Photos 14-28 Zip ]
[ Download Awards Ceremony Photos 29-43 Zip ]

Green Computing, Energy Efficiency and the Data Center

Avnet, Inc., Phoenix, Arizona – Case Study
FICO Corporation, Minneapolis, Minnesota
Infosys Technologies Limited, Bangalore, India
Lawrence Livermore National Laboratory, Livermore, California
MetLife, Troy, New York

Planning, Designing and Building a Strategic Storage Infrastructure

GlaxoSmithKline Biologicals, Wavre, Belgium – Case Study
Barclaycard US, Wilmington, Delaware
MetLife, Troy, New York
Sanborn, Colorado Springs, Colorado
TACO, Cranston, Rhode Island

Storage Resiliency, Data Protection and Recovery

FICO Corporation, St. Paul, Minnesota – Case Study
ABD Insurance (now Wells Fargo Insurance Services), San Francisco, California
Rogers Stirk Harbour + Partners (RSHP), San Francisco, California
Strand Associates, Madison, Wisconsin
Wentworth-Douglass Hospital, Dover, New Hampshire

Storage Virtualization and Cloud Computing

Medtronic, Inc., Minneapolis, Minnesota – Case Study
Budd Van Lines, Somerset, New Jersey
Citi, New York, New York
Cloud 10 Corporation, Centennial, Colorado
Service Corporation International (SCI), Houston, Texas

Technology Innovation and Promise

State Street Corporation, Boston, Massachusetts
Five Point Capital, San Diego, California
Marketing Architects, Minneapolis, Minnesota
Shopzillla, Los Angeles, California
Tickets.com, Costa Mesa, California

Special Thanks to our Judges!:

Rick Bauer, SNIA
Julia King, Computerworld
Ellen Lary, Lary.com
Richie Lary, Lary.com
Lucas Mearian, Computerworld
Brett Michalak, Tickets.com
Norman Owens, Carlson Companies
John Webster, Illuminata, Inc.
Laurence Whittaker, Hudson’s Bay
Ben Woo, IDC
Terry Yoshii, Intel

Eligible nominees are exclusively IT end-user organizations. Nominations of IT end-user organizations are accepted from IT users themselves, their public relations representatives, or vendors on behalf of their IT end-user customers. Multiple submissions of case studies describing different deployments per IT end-user/organization will be considered.

Virtualization Helps Avnet Avoid $5 Million in Build-Out Expenses

Avnet faced a significant challenge to its data center in 2005. It needed to expand its data center operations, but was already using 70 percent of its available power supply. Any expansions to its data center would require an investment of more than $5 million just to meet the initial power supply requirements.

As Avnet evaluated its data center environment, it determined that, like most data centers, a significant portion of its power supply and space were being used by servers and storage devices, while many of the devices were not being used to their full capacity. For example, Avnet’s x86 servers were utilizing 15% or less of capacity on average, while using 60-90% of their normal workload power even when idle. Additionally each of Avnet’s 900 servers accessed local disc and tape systems. The resulted in hundreds of drives and mounts at physical locations, all of which required significant power and cooling.

In searching for options, Avnet determined virtualization would help it overcome these challenges and provide a “green-focused” approach that would also benefit the environment. Over the next three years, Avnet focused on creating a greener and more energy-efficient data center, with virtualization serving as a foundation. This included:

Server Virtualization and Vertical Scaling. To improve its data center, Avnet focused on making its systems more efficient. It consolidated more than 300 pieces of hardware by retiring or collapsing systems into larger, vertically scaled servers. It also created a heavily virtualized environment, with 24 hosts supporting 450 virtual guests. Additionally, it deployed 15 product AIX frames, representing 538 systems across only 39 physical systems.

Storage Virtualization. Avnet moved from a local storage-based strategy with thousands of storage devices. It virtualized its storage in multi-tired SAN and NAS that allowed Avnet to properly size its storage requirements through thin provisioning. This approach enabled Avnet to realize economies of scale, while still maintaining the flexibility to allocate the right type of storage for specific identified needs. In total Avnet was able to consolidate more than 70 tape drives and over 100 locally attached storage units.

As a result, Avnet was able to avoid $5 million in build out expense to its data center due to server and storage virtualization. Additionally, it realized significant cost savings from the initiative. Server virtualization saved Avnet more than $1 million in hardware costs and approximately $250,000 annually in energy costs. It freed a considerable amount of raised flooring space within Avnet’s data center, which had previously been at a premium. Avnet now has 5,000 square feet of available raised flooring, up from 1,000 square feet. Additionally, storage virtualization has enabled greater power efficiency and decreased power consumption per server image by 44%.

GSK Biologicals Realizes Compliance by Design, Long-term Asset Optimization, High Integration and Simplicity

The GSK Biologicals headquarters is located on 2 campuses that are 10 miles apart in Wavre and Rixensart, Belgium. There is a data center on each campus; the heart of each is, by design, the SAN. SAN pairs are asynchronously replicated via two dedicated redundant metropolitan area networks. All production servers are connected to the primary SANs of the main data center, while all development and test servers are connected to the replicated SANs of the remote redundant data center. This low-cost design provides the highest level of compliance for business continuity, disaster recovery and record retention.

The IT department is operated as a “factory” meaning that each asset is optimized to its maximum potential. For the SAN, this means:

  • A strategic approach. The storage strategy was designed in 2000 and is now implemented, unchanged, during the 10 last years. The initial storage capacity of 20 TB then scaled up to a current 400 TB.
  • Continuous upgrades. Initial hardware has continuously upgraded in order to optimize performance, energy consumption and maintenance costs.
  • “Good is enough” approach. Over the decade, disk technology has dramatically evolved relative to performance and density leading to the coexistence of various models within a SAN. The objective is to avoid over-sizing and overperformance.

The highest level of integration is fundamental in modern IT to ultimately lower the TCO. In the GSK Biologicals headquarters, all major servers are connected to a single SAN, avoiding the hidden cost of alternative storage. In the 10 remote locations around the world, a standard architecture, including entry-level SAN, is currently deployed in “dark rooms” that are remotely operated.

By centralizing and virtualizing all storage, GSK Biologicals achieves a capacity utilization of 85% without jeopardizing the needs related to average data storage growth of 40% per year.

Key objectives are to simplify the landscape, avoid costly inter-operability challenges, and to make our technology partner fully accountable on the result.

By designing a low-cost, integrated, simplified, compliant SAN architecture, GSK Biologicals achieves an efficient cost per GB per year that includes hardware, software, DRP, backup, licenses, headcounts, and maintenance. According to a Gartner benchmark performed in July 2009 at GSK Biologicals, this cost is in the 25 percentile of the curve.

FICO Reduces Operating Expense by $1 Million While Improving Quality of Service

FICO maintains strategic data centers in both the United States and United Kingdom to serve its custom and corporate business functions. In early 2009, the CIO challenged the technology support teams to find innovative ways to reduce IT costs and consolidate hardware. One target for improvement was a backup service in a UK data center that was both problematic and expensive. The service was used to protect 8 terabytes of information across approximately 150 systems housing a variety of databases, servers and virtual guests. It utilized an undersized legacy 10/100 network that resulted in extended backup windows and recovery times for clients.

FICO support teams implemented a disk based backup and data deduplication technology. The solution included two appliances with 8 nodes and 12 terabytes of disk at each location. New clients were deployed to all 150 systems with an initial full backup created of all the data. The teams added redundant T1 connectivity between a UK data center and a FICO office to support cross-site replication between the new appliances. This provided not only traditional onsite data protection, but offsite replication of storage as well, alleviating the need to utilize offsite media management. The FICO solution also included daily reporting to the centralized storage team in the US to ensure all the client data was being successfully protected. The solution was so easy to use that the FICO Product Support Center was trained on new processes and procedures to execute backups and restores for customers.

With the new solution in place, backup media storage was reduced from 50 terabytes annually to less than 5 terabytes of deduplicated data on disk. The client based solution decreased backup duration, reduced bandwidth, and minimized host processing. Backups that previously took 2 hours are now completed in 15 minutes. Additional benefits included rapid file recovery from disk, elimination of offsite media storage and encryption of data at rest. This “lights out” solution could be remotely managed and allowed FICO to not only cancel the backup service, but also collapse a small data center to one of the local UK offices. Savings were realized with data center floor space, telecommunications, backup service fees, and offsite media storage. The overall reduction of the FY10 operating budget is expected to exceed $1 million. Management was excited to see the team accomplish the goal of reducing IT costs while also improving the quality of service provided to the clients.

Medtronic Finds Cost Savings of Virtualization to be Tremendous, and Agility to be Priceless

Medtronic currently supports over 2,700 Windows servers ranging from Windows 2000 through 2008 with four hardware vendors and 65 unique hardware types across 80 countries.

In just over two years, Medtronic has moved from negligible amounts of Windows server virtualization to fully 41% of the shared services environment being virtualized (approximately 1,100 virtual servers). Significant applications are now being virtualized including 100% of the Windows components of the SAP installation, portions of the Microsoft Exchange environment, and 100% of new sales force automation tools.

Medtronic’s capital expense savings as a result of virtualization are significant. In the primary data center alone, the estimated fiscal year 2009 operating expense savings is $9.6 million due to delayed growth of additional data center capacity. For Medtronic’s fiscal year 2010 (beginning in late April of 2009), Medtronic anticipates delaying the build out of an additional data center in part due to server virtualization. This results in capital expense savings of $30 million in fiscal year 2010.

As a result of Windows server virtualization, Medtronic no longer looks at servers in terms of physical rack units, but in terms of workloads that require a pre-defined amount of availability and computing power.