United we map: For GIS storage, bigger is better

A group of Western states wants to corral the costs of storing their geographic information system data, but they should first determine how large their posse should be.

Colorado, Montana, Oregon and Utah are looking to issue a request for proposals for a commercial cloud storage provider to host the states’ GIS data instead of each state negotiating its own separate contract.

But before pursuing a contract, the states should find out whether federal agencies might be interested in joining the consortium, a move that would considerably lower the cost of processing and storing GIS data for all members.

At least that was the recommendation of an advisory team the states formed to evaluate their plan.

State terabytes seek fed petabytes

As state geographic information system program officials shop for the best possible rates for cloud services, they are looking to bring federal agencies on board. That’s because federal GIS datasets are often more than 100 times the size of most state GIS datasets, enough data to significantly reduce shared storage costs. An informal search of published agency data found close to 26 petabytes of geospatial data at only three federal agencies, according to a multistate GIS cloud services assessment team.

Agriculture Department: 3.8 petabytes
U.S. Geological Survey: 4 petabytes
National Oceanic and Atmospheric Association: 18 petabytes

Source: Multistate GIS Cloud Services Assessment Team
(Colorado, Montana, Oregon and Utah)

Related coverage

A Western state GIS tour

GIS' future is with crowds, clouds ... and 4-D

In fall 2010, GSA awarded contracts to 12 vendors to provide cloud storage, virtual machines and Web hosting to federal, state, local and tribal agencies via the agency’s cloud storefront, Apps.gov.

In the absence of federal agency interest, a more thorough financial analysis should be performed using actual data volume and server usage provided by all states and major units of local government, the GIS cloud team recommended. 

The GIS cloud team's recommendations are based on input submitted through a request for information issued in November 2010. The CIOs of the states in the consortium are reviewing the assessment and recommendations.

Montana led the RFI effort, with active participation from the other states, which are members of the Western States Contracting Alliance. WSCA issued the RFI to assess the technical and financial feasibility of public hosting of GIS data and services that each state supports.

Montana accepted 23 RFI responses, ranging from prominent names in the industry, such as Amazon Web Services and Google, to some that were completely new to team members, such as Skygone. About 15 of the vendors could credibly participate in a formal acquisition process.

Strength in numbers

“If we can get enough states together to put their terabytes of data and some federal agencies with their petabytes of data [in the cloud], you can start driving down the costs of storage of GIS data pretty significantly,” said Dick Clark, Montana's CIO.

“Together, we have a tremendous impact into the marketplace; individually, we have almost no impact into the marketplace,” Clark said. 

The business model could help many states, especially smaller ones, with programs that receive appropriations funds from state legislatures. Most funding for IT shops around the nation comes out of states’ general funds, he said.

The federal geospatial community is aware of the Western states’ initiative, but no partnerships have been formed yet. The state CIOs still need to review the GIS cloud team's assessments, said Clark, who was the driving force behind getting the other states involved. 

“Basically, it is our GIS folks who are saying storage is expensive,” who want to find cheaper methods of storing GIS data, said Stephen Fletcher, Utah's CIO. The District of Columbia is also interested in joining the consortium, he added.

Moving GIS to the cloud goes beyond storage, said Robin Trenbeath, Montana's GIS officer. “We are talking about [computing] cycles — servers and cycles,” he said. 

Each state will need to deal with software on a state-by-state basis, he said. However, “when you get into IT units where you can share servers — that is really what we are talking about, at least at the beginning,” he said.

Mind the transaction costs

The states have three primary goals for considering GIS cloud services: managing costs, ensuring flexibility and scalability, and reducing staff support time.

Most states’ IT infrastructure is designed around transaction processing. The high-availability requirements of those types of applications might not be applicable to the full breadth of GIS processing. As a result, that environment could drive up GIS processing and data storage costs. The use of cloud services could avoid unnecessary processes and cut costs.

Cloud computing provides on-demand network access to a shared pool of configurable computing resources that can be rapidly provisioned and released with minimal management effort or service provider interaction.

Geospatial systems capture, store, analyze, manage and present datasets linked to locations. GIS is used everywhere, such as for protecting natural resources, providing public safety, redistricting political boundaries, drawing school districts, taxing entities and managing transportation, Fletcher said. 

“It is used many different ways, and as a result, there are hundreds of layers of data that are put on top of initial data and tied to specific points of geography,” Fletcher said, adding that the result is massive amounts of data to be stored.

The Western consortium’s GIS cloud team also noted that a government community cloud could aggregate the demand of all potential partners in the public sector, even some federal agencies whose rules might allow them to participate.

“That brings with it a governance challenge as well,” said Mike Boyer, Montana’s enterprise infrastructure project manager. Another recommendation is to set up a government GIS community cloud governance systems, he said.

“At the end of the day, this is really about proving the economics of the cloud [of what] states can do,” said Clark, who pointed out that states are currently confronting deep deficits. So it is important for all CIOs to look at their operations, reduce fixed costs and be more efficient.

GIS is just one example of commodity data that could be put in the cloud. In the future, state governments could aggregate other commodity data to lower fixed costs.

“That’s a win for the states,” he said. The states must have a win, Clark said, adding that “God only knows how we are going to get things across the nation funded in the near future.”