Networking Insurance Policy

Investments in networking, server and storage systems and a focus on continuity of operations can pay big dividends for state and local governments.

Nothing changes IT priorities quite like a natural disaster. Such an incident doesn't have to be catastrophic for agencies to put network and system upgrades that are part of continuity of operations on the front burner. 

For the Maryland Automobile Insurance Fund (MAIF), the weather emergency it encountered was a series of crippling storms in 2009 and 2010 that kept employees snowbound in their homes. The weather might have closed down MAIF's offices, but the work of the state agency, which offers liability coverage for citizens that would otherwise not be insured, needed to continue. Yet when too many of the agency's staffers tried to access the agency's servers remotely during the storms, the system crashed, and work ground to a halt.

The disaster was relatively small in this case. But during the snowstorms, "we realized that we needed to pay closer attention to how we respond when there is a disaster," says Matt Ailstock, MAIF's IT systems manager. 

Matt Ailstock of the Maryland Automobile Insurance Fund says the agency upgraded its network to improve its continuity of operations.

Photo: Gary Landsman

For Ailstock and his team, the ability to remotely provision new resources is critical. The 400-employee agency set out to rethink and redesign its continuity of operations and disaster recovery plans around a more scalable and virtualized data center environment. Although most of the agency's IT equipment resides at its Annapolis headquarters, MAIF also has two remote sites, one that supports the agency's legal team, and a disaster recovery site in Philadelphia.

MAIF needed what Ailstock called a "heavy-hitting" storage and server solution to support computationally intensive applications.  The agency chose server and SAN systems from HP and networking gear from Cisco, which were scalable and real-estate efficient because space in the disaster recovery site was limited.

Ailstock says the connection speed between the headquarters location and the disaster recovery site was a major component of the upgrade. The agency had a T1 line connecting the primary headquarters site to the disaster recovery facility. While the T1 was sufficient for moving small files, Ailstock says, it was inadequate for tasks such as replicating its Exchange database or its image-based claims applications. MAIF upgraded to a DS3 line.

"Speed has been one of the biggest additions to infrastructure recovery," Ailstock says.

The Cost Factor

Because cost was a major factor for MAIF, the agency was pleasantly surprised by how far its equipment budget went.

672
The number of simultaneous voice calls a DS3 line can carry

SOURCE: Gartner

"Maybe the economic downturn helped us [on pricing],"Ailstock says, adding that the agency executed the upgrade without having to spend money on consultants.

Although achieving a fast return on investment is top of mind for agencies, Lisa Erickson-Harris, research director at analyst firm Enterprise Management Associates, says there are other factors pushing governments to invest in new technology. One is a desire to ensure continuity of operations, another is a need to look over the long-term for potential benefits.

For the city of Mesa, Ariz., cutting costs was the prime catalyst in its decision to virtualize its datacenter. Faced with soaring power expenses and cooling costs and what was largely a hodgepodge of disparate data centers, the city consolidated its operations on Cisco's Unified Computing System. The Cisco gear supports computing, networking, storage and virtualization in a single platform.  

"The UCS solution seemed to fit our long term strategy best," says Joseph Sugihara, IT services leader for the city of Mesa, noting that in addition to offering streamlined and sophisticated management capabilities, at the time of the city's evaluation, UCS was the only blade solution that had support for converged architecture.  Since then, most manufacturers have added similar support.

The upgrade did require some significant changes to the city's network. Moving to a physical architecture that was collapsed onto a blade farm required the city to redesign its virtual LAN configurations in the datacenter to support collapsing multiple zones onto a single fabric, Sugihara says. "We also had to add fiber ports to the datacenter switches," he adds.

The results were impressive enough that Sugihara says if he had to do it again, he would make the same choice. Mesa slashed operating and equipment costs while expanding the citizen services it could support through technology, especially for its schools. The city entered into a partnership agreement with its school system to share the virtualized infrastructure for connectivity and radio communications. Mesa also plans to support video conferencing for schools through the infrastructure, opening up new opportunities for distance learning.

Lessons Learned

Here are some tips from Matt Ailstock of the Maryland Automobile Insurance Fund based on his experiences upgrading the agency's network for disaster recovery:

  • Focus on an upgrade that will correct an urgent problem, deliver a fast return on investment, or preferably both.
  • Get top management's buy-in from the beginning so that you can secure adequate personnel and other resources from the start.
  • Incorporate future scale requirements in the scope of the project.
  • Use manufacturers as a resource to scope out what's practical and what's possible.
  • Anticipate delays and build those into the project plan.
  • Allow for adequate testing time