The Circular Future - A Quantum Lifecycle Partners podcast

48. What Data Centre Decommissioning Means

• Quantum Lifecycle Partners • Season 1 • Episode 48

Share episode

Send us a text

The often-invisible backbone of our digital world, the physical infrastructure of data centers, carries significant environmental weight when it reaches the end of its operational life. This podcast episode delves into the crucial yet frequently overlooked process of data center decommissioning. Mark Petrie from Quantum Lifecycle Partners sheds light on the complexities involved as organizations increasingly migrate to the cloud due to the demands of artificial intelligence and the need for technology refreshes, all while navigating cybersecurity concerns. This shift generates a constant stream of retired IT equipment, laden with both valuable materials and sensitive information.

Key Takeaways:

  • Cloud migration and AI growth increase data center decommissioning, requiring secure and responsible handling.
  • Successful migration needs careful planning: application mapping, prioritized steps, and contingency plans.
  • Choose certified partners (NAID for security, R2v3 for environment) for responsible decommissioning.
  • Proper handling enables circular economy benefits: mineral recovery, reduced impact, and potential returns.
  • Explore resources like Quantum Lifecycle Partners for secure, sustainable technology transitions.


Thanks for listening!

If you like our podcasts, please leave us a review on Spotify or Apple or wherever you get your podcasts from.
Want to be a guest on The Circular Future podcast? Email Sanjay Trivedi at strivedi@quantumlifecycle.com


Speaker 1:

Data centers run the backbone of the internet, so how do we embrace circularity in data centers?

Speaker 1:

Welcome, to the circular future, your access to thought leaders and innovations to help you be a business sustainability champion, even if it's not your core job. I'm your host, stephanie McLarty, head of Sustainability at Quantum Lifecycle Partners, your trusted partner in electronics circularity. In our last episode, number 47, we talked about the environmental impact of artificial intelligence and we touched on the important role data centers play. Today, we're doing something a little bit differently and we're diving into this specific area of data centers with a short and sweet episode. What does circularity for data centers look like? What do we need to know when the servers and networking equipment get to the end of their life? With me is my colleague here at Quantum Lifecycle Partners, mark Petrie, who is our manager of business development for what's called SSN Servers, storage and Network, which is also known as data centers. Welcome to the podcast, mark.

Speaker 2:

Good morning. Thanks for having me, Stephanie.

Speaker 1:

Well excited to dive into this area and I have to admit we generally talk about electronics as a whole and the impact around circularity, basically anything with a battery or a plug. But today we're diving into the specific topic of data centers, which I'm not so versed in, but I know you are. I first want to start off this conversation looking at data centers and really around the decommissioning of them. Why do companies choose to decommission their data centers?

Speaker 2:

So probably the main reason people decommission equipment within their data centers is simply because they're doing a technology refresh. The existing platform has gotten a bit long in tooth and it can't keep up with the application or the growth in the application, or it's becoming too hard to maintain. So they just decided to refresh the technology and replace it with newer, faster, better performing technology. One of the other reasons they do it is because they've decided you know what I really don't want to be managing all these applications and all the associated hardware, so I'm going to move certain applications into the cloud. So they just migrate all of that application into the cloud and then they retire that equipment. And then probably the last reason for equipment is they've just the application itself has run its lifecycle and is now being retired, so they're going to let it lapse into obscurity and just replace that or remove that equipment.

Speaker 2:

Now there are some cases where customers decide you know what our core business is retail or pharmaceutical, or banking we really don't want to be in the IT business in which case they outsource everything in the data center to an outsourcing company, or perhaps they move everything into the cloud. At that point they're not just getting rid of equipment in the data center. If they decide to repurpose the data center, they're going to get rid of all the equipment and then all the infrastructure that supports the IT infrastructure. So that's the race floor itself overhead cabling trays, power distribution, cooling chillers, backup power. They may even have diesel power generators. So all of that stuff that would normally be associated with the data center is now removed. So those are the main reasons why people decide to get out of the data center business.

Speaker 1:

Right, and that makes a lot of sense. And what you've touched on, too, is that there's a lot more gear there than just the actual servers and the storage equipment. It's everything that runs that data center. You mentioned about moving to the cloud and outsourcing. Let's dive into that a little bit deeper. Why are companies doing that? Why are they moving from the traditional data center model to cloud-based models?

Speaker 2:

So at the beginning you touched on AI, and that's one of the big drivers right now, as people are moving to AI. But, as you know, it requires a huge amount of power in addition to regular processors, and because it draws more power, it also creates more heat. So then you've got more power requirements just to run the processor and then more power requirements for the cooling as well. So in some cases it can be a little difficult for customers, especially if they're in an environment where they don't have access to more power and they don't have access to more cooling and have access to more cooling. So rather than take on that challenge, they've just decided you know what I'm going to move to the cloud and let somebody else, whose business it is to run a data center, to handle that.

Speaker 2:

And then the other thing in my mind that plays into this is cybersecurity. Everybody right now is paranoid about getting hacked and getting held for ransom. So there's actually companies out there that will provide that as a service, but a lot of people are simply moving to the cloud because now that becomes the responsibility of the cloud provider. Yes, you still have to put certain things in place on your end, but by and large. The majority of the infrastructure to support your security is handled by somebody else, so you don't need to have experts in that technology and it's just easier for most people. In fact, a lot of smaller and medium-sized companies are choosing not to even have data centers. They're just going right to the cloud.

Speaker 1:

Wow yeah, cybersecurity is definitely a hot topic for everyone, and that makes a lot of sense. Let someone who specializes in this, who has experts in this, take care of the certain pieces of equipment, or whether you're doing a complete refresh or a complete decommissioning, you have to think about the business continuity side, to ensure that you don't lose any important access to documents, files, emails, even the internet, since we're all so dependent on it. What would you say are the things that a business needs to consider in terms of business continuity during a data center migration?

Speaker 2:

So there's a number of things. One of the first things you have to do is application mapping. And to explain that, I mean, let's say you have a web server that your customers log into and they select something that they're going to purchase from you, and then they go ahead and process all that. That web server could then transfer that data to another intermediate server which takes all that data and now makes it available to your ERP system, which processes the order, a CRM system, which is how you manage a relationship with the customer, and even potentially, a financial system. So now you've got potentially five different servers that are involved in this process, as well as the networking infrastructure and the storage behind it. So whenever you're moving something, you want to make sure you move all the systems associated with an application at the same time, so there's no disruption to the application. Then one of the other things you want to do is when it comes to scheduling. You want to rank your applications from most critical to least critical, and the reason you want to do that and it assumes that you have the time to move things individually you want to do the least critical things first. So if you've got to develop an environment or you've got a user acceptance testing environment or whatever it happens to be. You want to move that one first, and then what that allows you to do is to do a postmortem, so if there's any hiccups on the way, you can identify those hiccups and make sure that you smooth them out the next time around. It also allows all the people, from both the partner that you're working with as well as your own staff, to be familiar with what their roles and responsibilities are, so that the next time they do it it's better. So by the time you come to the most critical applications, you've done it a few times. I'm not going to say it's routine, because it's never routine, but at least you've got some experience with it, so that there's less stress on you when you do that.

Speaker 2:

The other thing I would suggest is that if you're contemplating a technology refresh, now is the perfect time to do it. You can stand up the net new environment at your new facility, get it up and running, replicate data from your existing site to the new site, run both systems in parallel and then, at such a time as you're comfortable, you can take the old system online and switch everything over to the new system, and I have customers that will run the old, the new system for a week or two just to make sure there's no hiccups, before they actually shut down the old equipment. And then again you don't have to move the old equipment because you've set up everything new in the new environment. And then the last thing is anytime you're doing something like this, you have to have a contingency or a fallback, which just means that you identify a person could be the application owner or whoever it is that if you've moved to the new facility and something's not going right at some point in time, there's got to be okay.

Speaker 2:

You know what it's either a go or no go, and somebody has to be selected to make that call and say you know what? We're four hours in. We've only got an eight-hour or ten-hour window. It's not working. Bring it all back, redo, you know. Bring the application back up at the old facility and we'll try again next week or in two weeks. So those are the main.

Speaker 1:

There's other things, but those are the main things to me that you need to consider when you're doing a migration from one data center to another. Well, it all makes logical sense, really, what you just described. Are you seeing that companies are making mistakes in this area? I would imagine, if this is something that they're in charge of, that they would know this, but perhaps that's wrong.

Speaker 2:

Perhaps we do see mistakes being made. What are the most common mistakes that you see In my experience? And it's they don't, I guess, account for time, and by that I think everybody's optimistic that they put a plan down on paper and everything's going to go according to plan. And I don't know how many times. I mean, we're talking to a customer right now that was supposed to start a migration as of January. It hasn't started yet.

Speaker 2:

It's looking like it's going to be May or June, and the reason for that is anytime you're doing something like this, you've got to get buy-off from a lot of different stakeholders. There's a lot of interdependency, so you have to have a project manager that's going to ride herd over this, because it can go sideways very easily. So you've got to have somebody that's making sure they're on top of this all the time. And you just have to assume that, and I keep saying I used to work with colleagues of mine and say, okay, the time I have a project, go completely without any hiccups, I'm going to retire, because that's a good way to retire, because you've got everything go perfectly, so just expect that something is going to happen. It could be minor, but just be aware that it's probably going to take a little longer than you expected or it's not going to start on time.

Speaker 1:

Right, so allow yourself extra time and sort of plan for the worst, but expect the best, kind of thing. Yes absolutely Love that. Okay, let's get into our how-to section, where we dive a little bit more into how to do things, where we dive a little bit more into how to do things. So, first of all, mark, how to choose a partner for your company's data center decommissioning needs. What should a company look for?

Speaker 2:

So I think the first thing you want to look for is somebody who has got experience doing that. Regardless of what it is you're looking at, whatever the project, you want somebody who's done it before, knows what they're doing and can handle it. In addition to that, there's a lot of regulatory compliance that goes around decommissioning stuff. So first you've got it from an information security standpoint, you want to have somebody who has NAID certification, and NAID is the National Association of Information Destruction and NAID is the National Association of Information Destruction and that is an organization that certifies you because you've done certain steps to qualify for that. And they also do inspections, unannounced inspections, where they will come in and check out your place of business. They'll also follow your trucks around to make sure that your guys are securing everything, locking everything, not leaving anything unattended. So that's one thing from a data security perspective. The other thing is if they have ISO 27001, that's an ISO standard, again with data security. It's similar to NADE and they're going to do similar type things to verify you.

Speaker 2:

Both of these certifications are things that are evaluated on an annual basis to make sure that you're current and you still pass. So that's one of the things I would look at. And then, from a environmental standpoint, you want to make sure they have certifications as well. So obviously there's the R2V3, which is the responsible recycling standard, which is not just affect the company that's doing it but also any partners they deal with downstream. So I mean if you're a company say a large bank or a publicly traded company, and you publish things on your annual report about stuff that you've recycled and you hand it off to company XYZ and they're not R2 certified and they just hand it off to somebody else who disposed of it in landfill, and if that ever came out, that doesn't look too good on you. So you want to make sure that whoever you're dealing with, not only do they handle the equipment and the goods properly, but when they pass it off downstream, that company is also R2 certified and will hand it responsibly.

Speaker 1:

Yeah, I know Quantum has both NAID, aaa and the R2V3 certifications and I know we are audited regularly recently finished our audit season, if you will and they are tough certifications but they make all the difference from a data security standpoint as well as environmental standpoint. Mark, what would you say how to get the most out of the data center equipment in terms of circularity, like how to reuse it, recycle it, anything? Any tips for companies there?

Speaker 2:

I think you want to look for a company that has a lot of downstream partners that they can deal with so that there is an opportunity to remarket either as an entire piece of equipment or at a component level within the equipment and, as well, they also deal with people that recycle equipment and can take it and then reuse the materials in a manufacturing process down the road. So that provides you with not only some money up front, if there is a remarketing opportunity, but it also ensures that none of this ends up in landfill and it's going to be reused. It also reduces our dependency on. You know they call the rare earth metals and there's a huge amount of energy that is required to mine and to distill so that you get the exact rare earth that you're looking for. So those are the main things that I would suggest.

Speaker 1:

Yeah, that's great feedback. So financial and environmentally, that makes sense. And yeah, we've actually just done an analysis at Quantum and we see 10 of the 34 critical minerals on the critical minerals list in Canada come through our processes at Quantum. So you're doing something great for the planet by ensuring your equipment is reused or recycled responsibly. Mark, this has been so insightful. I feel like I've learned a lot about data centers today. Let's finish off with one piece of advice which we ask everyone. If you could give one piece of advice in terms of data center decommissioning, what would you leave listeners with In terms of data?

Speaker 2:

center decommissioning, what would you leave listeners with? I think get a jump on it early, because if you are planning a decommissioning, especially if it's the entire data center, there's an awful lot involved in that and you've got to schedule in partners that are going to take out certain pieces of equipment. If you're taking out all the infrastructure, like the raised floor, that's a long process. It's probably going to be somebody different from usual, or you've got to find a company that will take all that equipment or all that material. So the sooner you start on this, give yourself plenty of lead time before you actually start it, just so you can make sure you've got everything all your ducks in a row. Because, as I said, there's a lot of moving parts, there's a lot of interdependency. So the more detailed planning and it goes for any project for that matter the more detailed planning up front, the less chances something slips through the cracks and the better you are to react to it if you do encounter a hiccup along the way.

Speaker 1:

Give yourself time. Yeah, one thing I've heard today is how this is similar to other types of electronics and technology that we would deal with at Quantum, but at the same time, there's so many more complexities here, given the interdependence with other systems and the cybersecurity aspect. So give yourself time. I think it's amazing advice. Thank you so much, mark. Really appreciate you having you here.

Speaker 2:

Thank you, pleasure being here.

Speaker 1:

And remember, if you're looking for a partner to help you repair, reuse and recycle your electronics, specifically data center equipment, we'd love to chat Head on over to quantumlifecyclecom and contact us. This is a Quantum Lifecycle podcast and the producer is Sanjay Trivedi. Thank you for being a Circular Future champion in your company and beyond. Logging off.

People on this episode

Contributor

Stephanie McLarty

Host
Contributor

Sanjay Trivedi

Producer