Always updating business technology trends
Get updates that impact your industry from our GigaOm Research Community
Can we enter the era of hands-free data centers, where remote software and robots handle jobs that still belong to human technicians to this day? That outlook may not be as far off as you think, according to one InformationWeek article John Edwards’ s drive explores the drive to make data centers autonomous. As reported by Edwards, the COVID-19 pandemic helped to solve the problem, with data centers around the world operating with a plummeting in employee numbers.
GigaOm analyst Ned Bellavance was quoted in the article. He urged IT managers to establish the right platform for automation efforts, warning that existing data center implementations could be difficult to transition to completely hands-free operations. He emphasized that a homogeneous and standardized environment is important for success.
As a case in point, Bellavance selected Microsoft’s Project Natick, an effort to develop closed data centers that could be deployed in coastal waters on the seabed. In 2018, Microsoft deployed a 240kW data center with 12 racks and 864 servers off the coast of Scotland as part of a Phase 2 test, as Bellavance quips:
“If you want to know something [a] turn off the lights really [data center] Looks like, take a look at Microsoft’s Project Natick. It’s quite difficult to send a technology on the seabed. “
Figure 1: Microsoft technicians slide a rack of data center servers and infrastructure into an undersea container for deployment to the seabed off the coast of Scotland. (Photo by Frank Betermin)
Leaving aside ambitious projects, Bellavance warns that achieving a hands-free, light-off datacenter is no small task.
“The reality is it’s very difficult to put all the necessary parts together for a data center to actually turn off the lights. You are looking at a lot of different systems that may have their own proprietary formats and protocols, ”he said.
Good news? Bellavance says progress is being made to establish useful standards, such as Redfish for managing servers, networking, and managing out-of-band power. These efforts are especially important, he explains, because a single tool can hardly manage every aspect of the data center.
“For that reason, I would be looking for management software that works well in a particular field and has API hooks for a orchestration class,” says Bellavance.
So how can IT organizations prepare for a daring future filled with flying cars and self-driving data centers? Bellavance, author of the recent GigaOm reports on edge infrastructure and edge placement design, offers a few tips:
- Choose a side: Either standardize on a single vendor and platform or embrace an open standard for management, says Bellavance.
- Skilled: Hone your automation and coordination skills, especially working with RESTful APIs.
- Small start: Start automating common tasks now and try to figure out how you can eliminate trips to the data center.
- Keep quantity: Make a list of common practice tasks and prioritize them by frequency and complexity.
- Admit: Invest in hardware with high levels of redundancy and low average failure times.
- Fail gracefully: Accept that failures will occur and plan to properly handle them through proper design and architecture.
- Consider AI: AIOps tools (see GigaOm Radar report) promise intelligent anomaly detection and even automatic response. Bellavance says it’s worth keeping an eye on these tools, but be wary of the great claims.