The promise of cloud computing as a cost saving effort hasn’t materialized like many expected over the past seven years. Sure, there are instances, particularly with email, where agencies are paying less for the same sets of services, or paying the same but getting more functionality.
But overall, few experts these days say the reason to move applications and systems to the cloud is only for the savings.
“We are not seeing huge cost savings in the immediate future because the only way we can save money immediately is to stop running legacy systems, shut down data centers or reduce their footprints,” said John Hale, the Defense Information Systems Agency’s chief of the cloud portfolio, at the May 3 FCW Cloud Summit in Washington.
. “We are not seeing huge cost savings. The reason why cloud got in the room was the potential cost savings, but now it’s because you get more agile, more functional capabilities.”
Hale said if DISA, for example, moves 990 out of 1,000 apps to the cloud, it would still have to support those 10 other apps in their own data center. And the cost to support those 10 apps, while maybe not as high as supporting 1,000 apps, still requires significant resources.
“If you could move those 10 apps to another data center and close that one, then you could save money. But that is a long-term reward and that is why moving workloads to cloud is not providing an immediate return on investment, which makes it difficult for CFO types to deal with,” he said.
Many agencies already have moved the easy stuff to the cloud, obtaining better services and limited cost savings or avoidance, and this is why the government’s move to cloud services actually is expected to decrease in 2017. IDC Government Insights predicts federal spending on cloud services to drop 17 percent to $2.2 billion. The decrease comes after several years of double-digit growth across agencies.
Shawn McCarthy, IDC Government Insights research director, said while the expected decline was surprising, it appears to be an anomaly.
IDC Government Insights predicts a return to significant growth through 2021, where spending is expected to reach $3.3 billion.
One way the agencies could spur growth in cloud sooner is improving the acquisition process.
Hale said the challenges of acquiring cloud services is the second biggest obstacle to moving a greater number of applications and systems to the cloud.
“We looked at options of how we could do that and worked with the Defense Working Capital Fund team and came up with this idea, and what I like to refer to it as the ‘Easy Pass model,’ where we set up an account, [the customer] puts money into the account in a Defense working capital fund situation, and then we have the vendor provide cloud services and actually take money out of the account like a utility-based billing model,” Hale said. “Much like how your Easy Pass works, when you get down to certain threshold, you get notified you need to add funds to it, and then you can add funds to it. If it ever gets to a situation where it runs down to a zero balance, then the vendor would stop your services. Vendors can’t end up in a situation where they are in an Anti-Deficiency Act (ADA) situation where they are providing services for free to the federal government.”
Hale said by having this revolving fund model and letting the vendor bill against it, DISA and its customers can ramp up or down the amount of computing, storage and other services its needs by the hour or day or week.
This concept is still in the early stages of testing. Hale said DISA is offering this approach through its MilCloud 1.0 services, which are all on-premise government-run clouds for DoD mission partners.
Hale said as DISA moves to MilCloud 2.0, where contractors offer access to cloud services, this model will become more important.
“There are four key applications that were chosen originally. One was an Army application and the other three are the Fourth Estate applications, including one that is DISA’s,” Hale said. “It’s in the early phases, but it will be something that will be widely available to the DoD in the next eight-to-10 months.”
Hale said DISA is finalizing its source selection process and expects to award the MilCloud 2.0 contract in June.
What maybe is most interesting about DISA’s approach is it’s taking existing payment models and applying them in a little bit of a different way.
“We did work very closely with the Federal Acquisition Regulations and Defense FAR in how the clauses are written versus the acquisition of cloud. The way we are doing this related to on-premise commercial cloud and doing the ‘Easy Pass’ model, it’s the same way the DoD has set up the Defense Working Capital Fund in the past.”
Hale said so far this approach is working well, but the true indication will come after military service and Defense agencies begin using the model to buy services through the MilCloud 2.0 vehicle.
DISA’s innovation comes at the perfect time, especially as Congress considers the Modernizing Government Technology (MGT) Act. Under the bill, agencies would set up working capital funds to help fund IT modernization efforts based on savings from moving to the cloud and other changes.
If DISA’s approach is successful, other agencies could follow suit to make moving the cloud that much easier, and move IDC Government Insights’ predictions to reality.