In the last week, the FAA grounded all flights nationwide due to a software communication issue. Millions of users couldn’t access their email due to a Microsoft Outlook outage. And customers of several financial institutions experienced a significant slowdown in their online banking.
In an era of cloud-based computing, even a small computer issue can snowball into a national shutdown.
To make sense of what’s happening, our Lucas Willard spoke with Ariel Pinto, Associate Dean for External Partnerships and Professor in the Department of Cybersecurity at the University at Albany’s College of Emergency Preparedness, Homeland Security and Cybersecurity.
Lucas asked if it’s accurate to describe these world-wide systems as “fragile.”
INTERVIEW TRANSCRIPT
Editor’s note: This conversation has been lightly edited for clarity.
Ariel Pinto: Being fragile, that's a great way to describe these systems. Meaning: they are very good, very reliable when they're working, but then we recognize it doesn't take a lot for them to start failing. And that is what we have heard happened in the Microsoft Outlook failure - is that there was this glitch in how the traffics are being managed from one server to the other, and that caused one server to fail, followed by another and another. And so, in that way, we might say that the server themselves were functioning as designed. But there is this layer of fragile software that failed to work at that moment, at that day, which has been working for so many years. But for one day, it did not, and that's what we had last week.
Lucas Willard: So, when you have so many people that are using one program, like Outlook, or when it comes to a banking software, and you have customers all over the world who are using one banking software. Is this sort of a necessary evil for having so many people using one system? What have the companies done to make sure that things are varied enough, or switched up enough, so that if you do have a failure in one system, it doesn't turn into what we've seen, which is this kind of chain reaction for servers all over the world?
Pinto: So, what we just experienced, whether it was with Microsoft Outlook or, the FAA, or the slowdown in banking transaction, this is what we call a systemic risk. Systemic meaning the interconnection of that system together with millions of users, makes a small failure affect so many other aspects of our society, far beyond simply the technical failure of a software. So that slowdown in banking transaction - that could have affected not just the account holders. If these account holders are businesses, then that might mean slowdown in their business transactions, which may mean that slowdown in the business transaction of their vendors and of their customers, and so on and so forth. So that is what makes the consequence of this failure so severe, because of what we call systemic risk, wherein one failure can affect so many parts of our society.
Willard: So, I guess that goes to my other question then: what can be done to stop these things from happening? Are there investments or ways that companies can actually set up their systems so we don't have one small problem snowball into this worldwide shutdown?
Pinto: Definitely, there are. Especially in the financial and banking institutions, they are required by regulatory policies to make sure that they have a robust system that can weather small failures, okay, small bugs in their software, that they can continue their operation. And if something really needs to be shut down, it would not mean the shutdown of the entire banking system. So, there are regulations, especially in the financial industries. But that doesn't guarantee that big incidents would not happen. That means that, aside from the financial industry trying to make sure that they can gracefully weather through any disruption, is that we, the consumer of their service, should also have some preparation in mind. If I am a business owner, and after this incident, I might start to think, well, “How can I prepare for the next time that my financial account is not available for one hour? Can I handle that? Are there alternatives for me to do to keep my operation going?” So, from the system owner's perspective, the owner of the banks or FAA or Microsoft, they already have within their company ways to make sure their operation is reliable. And in fact, they are very reliable. The mere fact that we are surprised by this happening meant that in recent memory we cannot recall the last time this happened, and so which meant that they are reliable. But also proof that they can still happen, which is a reminder for us, the larger society, to ask ourselves, “Well, what's my plan B? How can I weather away a few hours of not having an email or not have access to my financial accounts?” So those are on our side.