Virtualization Technology News and Information
Why decentralized work calls for decentralized data

By Jack Coates, Senior Director of Product Management at Tanium

Prior to the COVID-19 pandemic, I spoke with a company that was collecting several hundred gigs of highly granular CPU performance metrics, from every endpoint, per day. While this seemed incredibly inefficient and wasteful I was told, "bandwidth and storage aren't that expensive."

We know what happened next. Millions of endpoints moved to residential broadband networks, and bandwidth suddenly became a real challenge.

As organizations emerge from pandemic "survival mode" and look toward long-term growth, they need to re-examine how they manage their corporate IT. With millions of people now accessing networks from home, sometimes from personal or unvetted devices, the number of edge devices - and corresponding data - has grown dramatically. In the final quarter of 2020, consumers worldwide snapped up more than 385 million smartphones and 80 million PCs, up nearly 11% from the previous year. Unfortunately, the way enterprises manage these devices hasn't kept up.

The traditional data lake approach, which operates on the assumption that most end-user devices sit behind the corporate firewall and can be managed in a central location, lost relevance well before the pandemic but was rendered obsolete almost overnight with the shift to remote work.

Because organizations are dependent on people creating and using data across increasingly distributed networks, the most efficient solution is to instead execute at the endpoints where those people actually work.

There are several benefits of this decentralized approach that keeps data at the edge, including:

  • Speed. Consider a typical security function, such as a domain administrator logging into a known-compromised endpoint. Using a traditional data lake approach in a modern distributed enterprise requires moving login and endpoint records into storage, normalizing these records, and detecting a match, a process that can take up to 45 minutes. Producing that same answer at the edge as little as 15 seconds.
  • Cost savings. Data lakes require significant infrastructure, whether purchased as a service or built on-premises. By asking smart questions of small slices of endpoint resources, decentralized IT management requires far fewer resources than centralized approaches - and this is an increasingly critical priority among overtaxed mid-size organizations.
  • Reduced risk. All the metrics and events in the world are useless if they're inaccessibly buried in a data lake. By teaching the endpoint to report on what matters, organizations can save resources for higher priority, higher risk events.

To transition to a more modern approach that saves time and money, and lowers the corporate risk profile, organizations should move to a decentralized device management mode. But how can this be accomplished?

1.  Get buy-in

Any organizational shift requires strong alliances and agreement throughout the organization. Most importantly, business leaders need to feel that there is demonstrable benefit to the business, whether through saved costs, increased productivity or other measurable ROI. IT leaders looking to move to a decentralized device management model should come prepared to make the case by understanding the costs, in terms of raw dollars and human labor, as well as the limitations of the traditional approaches for their specific organization. They should also have a clear road map that shows how the organization makes the transition, and how the new paradigm will sustain momentum once achieved.

2.  Source the right tools and tech

The right tech is essential to the long-term success of a decentralized management model. When mapping out the transition and long-term plan, IT leaders should always keep the end goal in mind: speed, cost savings and lower risk profile. Most importantly, the tools must be fast, effective and flexible to prove useful. A product that excels at one function, but can't be used for anything else, is tough to justify. Likewise, another that can theoretically do anything, but has no out-of-the-box functionality, won't suffice. Edge tools should be able to answer questions and provide actionable intelligence immediately, without locking teams into a few fixed use cases.

3.  Bring the people along with you

Shifting to a decentralized IT management model requires not only a change to your organization's data architecture, but also a shift in the way people engage with the architecture, the skills they bring, and their willingness to change. With any paradigm shift, including to the IT business, you have to manage the human side as much as, if not more than, the technical side. Small, quick wins and demonstrable outcomes are key to success.

With millions of people working from home in 2021, it's never been more important to manage, monitor and secure edge devices. Given today's situation, yesterday's approaches fall short. Keeping data at the edge allows it to tell you what's important, and permits you to do something about it, while saving time and cost and reducing the organizational risk profile. This in turn enables the organization to better serve its customers, and convert savings into investment in innovation and long-term stability.



Jack Coates 

Jack Coates is senior director of product management at Tanium Inc. Prior to joining Tanium in 2018, Jack has held senior positions at organizations including Splunk, BigFix, and LANDesk. He also holds four U.S. patents in machine learning, data visualization, and cybersecurity.

Published Friday, May 21, 2021 7:35 AM by David Marshall
Filed under:
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<May 2021>