Every IT team has been there—a data center shutdown leaves you scrambling to explain what’s going on to your customers and ensure you retain their trust.
Outages are serious business. A 2023 New Relic survey indicated that the majority of data center shutdowns cost upward of $100,000 an hour. And with potentially trillions to be invested in getting the centers up to speed for AI, data centers are more important than ever.
The list of reasons for a data center shutdown is long, F5 CTO of Government Solutions Bill Church told IT Brew. It could be maintenance, lease termination, or a number of other unscheduled—and it’s often unscheduled—circumstances, but the end result is the same: You’ll need to make sure that you’re covered, and hopefully before the inevitable occurs.
“Ideally we’ve documented everything, and we know what our resources are, who owns them, so we’ll go with that assumption, because there’s always corner cases where there’s stuff that’s missed or left out,” Church said. “But…if it’s completely unscheduled so there’s no warning, and there is no redundancy for alternate data centers or disaster recovery, your options could be somewhat limited.”
Good outreach. The key is to have good communication. Social media can help, Church said, as well as, if necessary, the press.
Joe Minarik, COO at DataBank, a data center management company, said that telling the customer about what kind of shutdown they’re experiencing—it’s most often either power outage or cooling related—helps to maintain that trust.
Top insights for IT pros
From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.
“In both of those key aspects, there’s customer communication to help the customers understand what is happening,” Minarik said.
People “freak out” when their businesses and livelihoods are on the line, Minarik added, and often that’s the case with data centers. Sending out frequent notifications is helpful to let customers know where things stand so they, in turn, can pass that information on to their customers or users.
There’s a chance. While the situation isn’t ideal, a shutdown could offer unexpected opportunities. You can rehome or rehost apps, move things to the cloud, and look into other services.
For companies like DataBank, the top priority is to get things back up and running—the company promises 100% uptime, Minarik told IT Brew. But if given the chance, they’re able to use the downtime to take care of certain things that need doing. Minarik stressed that’s a rare occurrence.
Church referred to the CrowdStrike outage in July 2024 as an example of learning from mistakes. The “blue screen of death” was a way to have insight into “a neglected part of people’s environments” like kiosks and displays. That the fix necessitated human intervention showed what shouldn’t happen moving forward, and, to Church, the same applies when data centers go down.
“It’s a good opportunity to look at your disaster recovery plan and see what has not worked, what is working,” Church said. “And it’s a great way to test that, because it’s real, it’s happening.”