Thursday, November 7, 2024

Intel’s Jeff Klaus: Edge Computing and Data Center Management

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.

Clearly, managing a data center is much harder than it used to be. Emerging technologies like artificial intelligence and Big Data and IoT have increased the workload, boosting expense and complexity.

To provide guidance on these trends, I spoke with Jeff Klaus, General Manager, Data Center Software at Intel. We discussed:

  • Why it’s harder to run a data center than ever before.
  • The tools and solutions that can help the efficiency of data center management. The role of DCIM (data center infrastructure management) tools.
  • The mega trend toward collecting and processing data at the edge. While more than 90 percent of data is processed in data centers today, Gartner predicts that by 2022 about 75 percent will be handled at the edge.
  • Key issues and challenges. How can IT managers prepare this move to the edge?

Scroll down below the video to see highlights

Key Points: Data Center Management

Below are highlights from Jeff Klaus’s discussion of data center management:

Increased Complexity

“I give [data center managers] a lot of credit for managing the evolving set of technologies. We had this Big Data phenomenon and now it’s AI, and then you’re moving to the edge. And at the same time, the level of interconnects and devices that are connecting to the data center and the speed requirements – it’s a significant challenge.”

Managing Multiple Environments

“We asked [data center managers] how many remote environments that they’re managing and end points are they managing. And it turned out that close to 60% are managing five plus more, five or more unique environment.

“And it might be the traditional data center that we think of, but it can also be this phenomenon of edge and the continued movement of getting compute resources to the customer closer to the customer and more inexpensive positioning with the customer. And that level of complexity is going to continue, but it certainly has contributed to the challenges of data center operators today.”

Need for More Analysis Tools

“So we [at Intel] are trying to feed additional data to the operators so that they can make more intelligent decisions with this disparate remote environment they’re managing.

And what we’ve seen is that they’re asking us for more analysis tools. We went through this period of time where just getting the data was a struggle, and even in the survey that we commissioned, there were still about 40% of data center operators that are struggling to just get the data they need to manage their environment.”

Firmware Adding Complexity

“Just from talking to customers and understanding some of the complexity, we have a lot of revisions that occur at the OEM space and Intel contributes to some of that complexity because we self-obsolete ourselves every two and a half years or so with a new chipset. And that goes out to all of the OEMs and then they release new servers and new technology to their customers.

And what happens is, the OEM takes a lot of the baseline components that Intel is providing, and they add their firmware layers on top of the chipset or on top of the architecture, and the customization in firmware is really causing some complexity.”

Data Centers and the Edge

“The edge is faster and it’s cheaper. It’s closer to the customer. And depending on the type of services that are required, it’s a requirement to ensure that the customer’s information is processed that much more quickly in a customer type setting, rather than going back to a traditional data center.

“So the challenge is more remoteness. We talked before about five-plus remote environments, you’re going to see that significantly increase.”

The Edge and Analysis Tools

“So there’s generally not going to be a human being there to be able to remediate, fix, turn off and on, or analyze a set of hardware. So you need a lighter tool to remotely establish a link to find out or look and discover what type of issues could be occurring there.

“But you also don’t want a network hog or an analysis tool that requires a lot of network bandwidth or requires a significant amount of people to manage.

“So I think the toolsets, the traditional toolsets, and DCIM, have evolved into a set of buckets that are underneath that larger umbrella that are really just defining customer problems and addressing them. And Edge … has similar issues, but it’s just on a smaller scale. And what we’re seeing is, we see a lot of requests for analysis tools ‘I got all this data, but I want to understand how to interpret that information and what to do with it.’”

Handling the Proliferation of Data Center Tools

“I think there are many tools out there. I think that’s one of the bigger challenges, that I think data center managers are being hit up to evaluate something almost weekly.

“There are customers or partners that are in this space that are doing a good job on the business development by getting into the market and trying to grow. And it’s pretty easy to set up a software tool that can collect some information and evaluate it.

“But we’ve often advised customers to do a real diligent PoC. We were involved in Intel IT’s evaluation of an IT management, data center management system, and it’s a painful process.

“The IT department discovered [that] really understanding what it’s going take to implement and maintain a solution that really says it’s going to do everything for you, and that’s part of the promises that are made, [requires you to] set aside individuals and resources. Not only to implement, but also to have a much higher level of maintenance resources than you traditionally would believe.

“So, kicking the tires, doing some good diligence, getting some customer referrals, those types of really basic requirements are something that I would encourage all data center managers to do.”

IoT and Data Center Use Cases

“Then with newer generations, when you look at the evolution of IoT, its sensors are getting smaller and smaller to be able to put inside industrial equipment.

“Well, that’s essentially what’s happened within the IT devices: now there are more sensors that are within your IT devices to help monitor the health, monitor the temperature, and monitor the power utilization.

“So that has blossomed into a whole number of use cases from this information. And how we’ve packaged our sets of tools is, ‘Tell me your top three problems, I have a portfolio of roughly 10 use cases that I can apply to your issues. Let me prove one of those use cases out to you before you make an investment in people or an investment in capital towards the tool.’”

Subscribe to Data Insider

Learn the latest news and best practices about data science, big data analytics, artificial intelligence, data security, and more.

Similar articles

Get the Free Newsletter!

Subscribe to Data Insider for top news, trends & analysis

Latest Articles