The Defense Department might be shooting itself in the foot in its haste to transition to the cloud.
The most recent example of DoD’s push to cloud adoption is the Defense Information Systems Agency’s request to industry for ideas for how to rapidly migrate to cloud services for MilCloud 2.0, which the agency released on Wednesday.
While few would argue DoD’s move to embrace a comprehensive, easily accessed system for information is a bad thing, William Schneider, a Hudson senior fellow and former chairman of the Defense Science Board, said DoD’s lack of strategy is hurting its cloud procurement. The event was sponsored by Microsoft and Oracle.
This is especially true for the most contentious and talked about cloud contract of the year—the Joint Enterprise Defense Infrastructure (JEDI) program.
Schneider said its implications of JEDI will reverberate for years to come because of the size, scope and cost of the $10 billion cloud contract.
JEDI “has become a little more contentious simply because it’s the first lurch into the cloud for DoD, aiming at a systemwide thing analogous to intelligence community’s move in 2013 into the cloud and it is a big change,” Schneider said in a Thursday meeting with reporters in Washington at the Hudson Institute.
Schneider said the contract will set precedence for future large contracts as DoD continues to adopt cloud services down the road.
And DoD already is botching it, Schneider said, by not laying out clear plans for the public and industry.
“Because DoD doesn’t have a roadmap, or at least they haven’t published one, about how they are going to get there, its animated all the anxieties about what is going to be the future of this market,” Schneider said. “Is it going to be served forever by a single-source incumbent like anchor chains where we have one incumbent that makes anchor chains and they’ve been doing it for decades or is this going to replicate the commercial model? DoD says they want to take advantage of the commercial model, but the absence of a roadmap has made it difficult to persuade the Hill, who tend to be the biggest advocates for competitive sourcing.”
Schneider said DoD needs some way to express where it’s headed, much in the way former Director of National Intelligence James Clapper did when laying out ICITE.
The intelligence community was “pretty clear as to where it wanted to go, but DoD has not gotten there yet in terms of explaining where it’s going,” Schneider said.
That is clear in the MilCloud contract as DISA works to migrate to MilCloud 2.0 and shutdown 1.0 by late next year, as DoD does its own procurement for JEDI.
JEDI has been a notoriously opaque contract. DoD plans to award the task order procurement to one vendor. Many in industry and Congress questioned that approach, but weren’t satisfied with DoD’s reasoning behind the strategy after the second release of its draft RFP.
Schneider said what’s most important in awarding cloud contracts is they are competitively sourced, provide a secure cloud and are innovative.
Despite objections from industry and DoD giving its chief information officer Dana Deasy control over the project, DoD released the third and final RFP in July still as a single award contract.
“I am confident the JEDI Cloud RFP reflects the department’s unique and critical needs and employs the best standards of competitive pricing, innovation, and security,” Dana Deasy, wrote in a letter to industry. “I am excited to be part of an initiative that will revolutionize how we fight and win wars.”
Schneider said he thought the final RFP was more of a restatement than an update to the RFP.
Oracle filed a pre-award protest on Aug. 6. Oracle filed a second complaint last week based on new amendments to the JEDI contract.
Despite spats over the sourcing, Deputy Defense Secretary Patrick Shanahan said DoD hopes to eventually award multiple contracts to different companies to move its IT systems to the cloud.
“When people say, ‘Oh, it’s a winner-take-all,’ it’s really something that’s less than 20 percent,” Shanahan told reporters in April. “People say, ‘Why would you pick something that’s less than 20 percent?’ Well, if I do two percent, there’s no real scale. If you do between 10 percent and 20 percent, you start to get some scale.”