Climate change is already starting to affect the world in cataclysmic ways. Climate science from the likes of the World Resources Institute gives us a generation or two to radically change our ways before the impacts become unalterable and serious on a systemic level in terms of human life as we’ve known it for the last machine-enabled hundred and fifty years. That has led to a real focus on pushing down our carbon footprints on an industry-wide level to achieve “Net Zero.”
In Part 1 of this article, we spoke to Paul Mackay, Cloud Director, EMEA at Cloudera, to ask whether getting to the notional “Net Zero” target – a state of overall carbon neutrality – was possible by 2050 (a goal set by the UK government).
Paul seemed optimistic that advances in technology and processes meant it was at least on some level feasible – though he highlighted the fact that it was going to be more difficult for public sector organizations to achieve than it was for private sector businesses.
Profitable zero?
While we had Paul in the chair, we asked him about the cold, hard business of cash.
THQ:
We’ve spoken to people in the data center world recently, and encountered the competing interests of planet vs profitability. Is that going to be a factor in hitting Net Zero targets?
PMcK:
Spend is a question that will absolutely come into it, yes.
I remember previous organizations where we’ve responded to an RFP and we’ve talked about sustainability. In some of those organizations, there was a recognition that they may need to spend more money in order to drive that agenda, and that by buying a more carbon-neutral or energy efficient or Net Zero-friendly technology, there’s an associated uplift in cost in that that may also be prohibitive for instance to a public sector organization that is working on even tighter budgets and is severely restricted by that.
Virtual zero?
THQ:
Let’s talk technology. In fact, let’s talk virtualization. How can virtualization help get us closer to Net Zero?
PMcK:
Virtualization has been around for a significant amount of time now – 15-20 years.
I remember walking through London when it was snowing. And if you walked past some of the banks, you’d see snow everywhere, except for on the paving stones outside the banks. That was because the data center was underneath the building, and the heat coming up from that data center would melt the snow.
Virtualization in its initial value proposition was aiming to solve that sort of problem – we don’t want physical servers losing all that energy, so let’s collapse those on to fewer servers, but use software to segregate them so that you get the look and feel of the physical server, but it’s virtual. And we saw this huge adoption. As organizations reduced rackspace, they started to reduce floor space in their data center and become more efficient with their hardware that allowed them to do that.
It made a massive impact. What we’re doing now with containers takes that even further, breaking down workloads and applications even further.
One of the problems with virtualization though was that all of a sudden it became really easy to spin stuff up. So whereas previously, if I had any projects, I’d have to go out on procurement and buy my hardware and rack and stack it and so on, now, all of a sudden, I could just spin up a VM, and have access and start to deploy software on top of it.
And so we saw VM sprawl. And I remember as I transitioned out of being on the technological side of things into a leadership role, that was becoming a big thing. Help reduce VM sprawl, because VM sprawl was causing those data centers that you’ve stripped down to grow back up.
And so virtualization as a technology, plus the introduction of cloud technology and how I can then break that down even further and start to do things like containers or start to do things that are microservices-based, plus the ability we have now to instantly go on and get access to it – that’s absolutely a fundamental way of reducing our traditional on-prem footprint.
I will say, though, that you can’t virtualize everything. There is stuff that that data center that I talked about, with the melted snow, still has to do. In the case of big banks, their systems are 40 years old. There are mainframes in there that are still functioning – when we log on to our internet banking app, at some point, we’re still talking to a mainframe that will always need to remain in that data center, because it’s too risky to move it, and it’s too expensive to move it.
A zero balance?
So you have to find the balance between virtualization, physical infrastructure, and cloud. And, you know, we see this shift now as organizations move to hybrid. They’re all wondering how they live in these different environments, in this combination of technologies, and get to a point where they’re running as efficiently as possible.
And then I’d add to my earlier point, when you start to add on more automation, you can actually start to reduce human interaction and start to do things in an even faster, more scalable way, which absolutely has a knock-on effect in terms of your carbon footprint, or the amount of energy you’re using.
In Part 3 of this article, we’ll talk about performance economics, and the role that hyperscalers can play in the fight towards reaching Net Zero targets.
