How to make a joint IT project work

IT staff at Bowdoin College, in Maine, and Loyola Marymount University, in California, are building a joint disaster-recovery facility, each institution hosting a disaster-recovery hotsite for the other.

The CIOs of both schools, Bowdoin’s Mitch Davis and LMU’s Erin Griffin, are insistent that their experience can be a model for other cooperative IT projects, even among business rivals.

The first step is to conceive the possibilities of cooperative projects, Davis says. “When Erin and I talked about this to [a gathering of] mid-market businesses, they said ‘we can’t do this: we’re in manufacturing,’” he recalls. But Davis pointed out they all buy materials from the same group of suppliers. Why not have the respective IT groups cooperate on a business-to-business purchasing solution and also back each other up, he asked.

“When you approach this with the sense that you can not do it, you probably won’t do it,” Davis says.

Be prepared for skepticism at all levels. “Conceptually, it so pushed the boundaries of everything we’d read in the literature about how to do things like this,” says Dan Cooke, director of systems administration at LMU. “It seemed really daunting when we first started.”

For both CIOs, the dominant attitude seems to be “we can do this, so get it done.”

“You can put a lot of people in a room and they can talk about getting something done,” Davis says. “But if you put a smaller number of people in the room and tell them to get something done, they have to act.”

Bringing the two teams together in such a context was an eye-opener. What happens is that the protective bubble around each group gets broken. “Outside of that bubble, someone can say to you, ‘why are you doing that?’” Davis says. “You can look at the results people are getting, find out what are the pitfalls, what they’ve learned. And you can take all that and apply it, and make that learning part of this new project.”

“As we worked [together], it started seeming much more tangible and realistic,” Cooke says.

“[The two teams] have a lot in common personality-wise,” Griffin says. “I think it’s been a lot of fun for them.”

Focusing on VMware and other technologies

Griffin describes the basic evolution of the project as an almost matter-of-fact process that competent IT professionals are well-equipped to handle. “We started by laying out on the table ideas about potential technologies,” she says. The team quickly decided to leverage the in-progress changes to both schools’ network infrastructure, specifically major network upgrades, a high-bandwidth link to Internet2, and especially server virtualization based on VMware software and blade servers.

“So we said, ‘let’s build [the disaster-recovery site] on VMware,” Griffin says. “Then we asked questions like, ‘how would that work? How can we leverage that?’” The teams decided what initial services and applications would be deployed in the disaster-recovery site. “We asked questions like ‘what do we need in place to build out these capabilities?’”