>> Narrator: From theCUBE Studios in Palo Alto in Boston, connecting you with thought leaders all around the world, this is a CUBE Conversation. >> Hi, I'm Stu Miniman and this is a CUBE Conversation, I'm coming to from our Boston area studio. Happy to welcome to the program, Angela Medina. She is the director of product marketing at ThousandEyes. Thank you so much for joining us. >> Thanks for having me Stu. >> All right, so Angelique, we get to dig into some research, a new report, it's set up to be annual, the 2020 Internet Performance Report. Of course, internet, like everything else in 2020, things are a little bit different. So help us understand a little bit the purpose of this report and what led to its inaugural incarnation. >> Yeah, absolutely. So it's really interesting. So about the March period, when suddenly there were shutdowns in the US and other regions, and a lot of the workforce was working from home, we started to get a lot of inbounds from our customers and other interested parties about how the internet was holding up. So there was systemic network degradation, are the operators able to handle all of this traffic and these pretty significant traffic shifts. So in responding to this, we decided to put out or make public our outage detection capabilities. So we put out an outage hub and this was back in May, and it basically shows ongoing and recent disruptions, so that's overlaid on a map and you can see where outages are taking place and which networks they're taking place in. So, that's been out for a few months, but we wanted to look at not just outages in real time, but historically as well. So, in looking at the last few months from January through the end of June, that's a really interesting time capsule because we can look at how networks perform and behave, not just under normal conditions, so maybe January and February, but also highly abnormal conditions. So it's a very interesting way to understand how the different types of network operators perform, given what we're enduring right now. >> Yeah, It's fascinating to me if you've watched the networking industry Angelique, normally these kinds of changes in networking are things that we measure in years if not a decade. I remember it felt like it was at least 10 years we talked about the mega trend of North-South to going East-West, how virtualization was changing everything. And of course in 2020, it's all of a sudden, all right, everybody work from home, everybody's home internet is going to be stressed. So help us understand a little bit how much of that is a blip that we saw over a very short period of time and what the resulting output is. >> Yeah, it was really interesting because beginning in... I mean, there are always a certain level of disruptions. Disruptions are just a part of operating a network. Occasionally you're going to have a little bit of downtime, stuff breaks, but in March we saw a pretty dramatic spike, particularly in North America and Asia pack. The level of disruptions, the duration of the disruptions and the scope of them. So more infrastructure impacted was pretty significant. So it was something like a 66% increase globally. And this did start to go down in subsequent months, so we're at a point where it's not quite back to January, February baseline, but it's pretty close. So we definitely saw an increase, but it seems to have stabilized as in a lot of areas, traffic has plateaued or normalized. >> And when you talk about the internet, of course, the internet is made up of lots of devices and lots of companies. What particularly is interesting there, if you look at some of the internet service providers out there, if you're look at companies that are doing remote contact centers, are you able to see a heat map or some of the areas that might have been under more stresses and strain? >> Yeah, absolutely. So we're looking at it not only from a geographic standpoint, but we're also looking at different network types. So we're defining the internet fairly broadly, not to just be connectivity as a service providers like your broadband providers or your transit providers, but also cloud networks networks. Cloud networks these days are really an extension of the internet. So we're also looking at their performance and how they held up as well as the networks of really key services like CDN providers, content delivery network providers, as well as DNS. So the collection of these networks is really what is the foundation of what most people think of as the internet. >> Yeah, it's the thing that we've been saying for a number of years is if traditionally you were somebody that managed the network, it used to be something that you would touch. And now, of course, most network operators, you are responsible for a lot of things that you don't necessarily touch. I'll give the disclaimer of course, anybody watching, Cisco has made the announcement to acquire ThousandEyes of course, the gorilla in the networking world. We look forward to talking about that more once the deal is completely closed. So Angelique, how are the cloud providers doing? How are end customers reliance on all of these various services? How are things holding up? >> Yeah, I mean, you bring up a really good point about the fact that a lot of enterprises have dramatically transformed or are in the process of transforming themselves where they're now so dependent on external services like cloud providers, like provider networks, more internet service providers, as opposed to managed services. So overall, there hasn't been any kind of systemic breakage across all of these providers, but really the devil is in the details. Oftentimes you might have issues where there might be of an increase in disruptions, or you can't quite pinpoint where the source of an issue is. So really being able to see into these external services and understand how they're performing and have that visibility so you can communicate is really important if you're going to be successful operating in this new IT reality. But in terms of how the different providers perform, how providers are vastly more reliable than internet service providers, probably for a number of reasons having to do with how they've built their networks. They're software defined, they're not as dependent on the underlying infrastructure. So they have much newer networks too, less technical debt, for sure. >> Anything specific when you look at the data over time, are we through the biggest shift in what's happening in more of the ripples now, or have things settled out a little bit, I guess, since some of the initial shocks? >> Yeah, so it varies by provider and region. If just think about the United States or North America. So in North America, we definitely have seen that the number of disruptions have come down over the last couple of months. And we're at a point we're really only about 20% off from what we were seeing in January and February, but cloud provider disruptions haven't quite returned to earlier levels. So they're still on the upswing. So it will be interesting to see where that goes, if that continues, or if that eventually starts to plateau and then decline. But they're going in different directions in North America, disruptions are up, but ISP disruptions are down. >> All right, maybe, could you explain a little bit, what does it mean by an outage? Actually, I pull up right now, the internet outage map which you have on your website and there's these scary, glowing red circles, but you and I are connecting from across the country, obviously using the internet to be able to do videos. So just because there's red glowing lights doesn't mean that we don't have internet access. >> Right, right. I mean, so just in terms of what an outage is, an outage, as we define it, is where there is a hundred percent packet loss within a provider's networks, so traffic is effectively terminating at an interface within the infrastructure of a provider. And so when that happens, we'll flag it. And this is based by the way, on billions of network probes that are sent over the internet using our platform each day. That's how we effectively derive these measurements. And when we see these disruptions, again, we'll flag them and to your point, yes, big, glowing outages on a map, but you know, you're right, not all of them were necessarily going to be disruptive to users for a number of reasons. One of the earlier points you brought up is that the internet is made up of thousands of independently operated networks. It's not like a utility, so you may go through a provider that's having an issue when you may not and that can change dynamically depending on where you're connecting to and what service you're trying to reach. It gets very complicated. >> Is that so? Yeah, so I'm curious one of the biggest challenges out there is that companies have to rapidly make changes. Whether it is adopting cloud services or getting ready remote call centers or the like, are there anything that they can take from the survey data or these maps as to how should I plan things? How should I make these changes? What can practitioners learn from this? >> So I think it's important to understand how do operators, what are their habits and practices, depending on where you're located and we've seen regional differences. So for example, in the US, with internet service providers, they tend to have disruptions that are taking place outside of traditional business hours. So less disruptive, more likely to be due to a maintenance window, change that they're making, versus other regions where many more of these disruptions are taking place at times that might impact a business. So understanding how different providers vary in terms of their practices, gives you an opportunity to have that conversation with providers to hold them accountable and to work collaboratively with them so that you understand when are they going to be making changes. If there are increases in traffic, maybe you have some resiliency measures in place because you know that the operators might be a little bit stressed responding to these increases in either traffic or changing traffic patterns. >> All right, are there any other key takeaways that people should take from this new report? >> Yeah, I mean, I think, one of the key things is that not every outage is created equal. Not all providers are created equal. I mean, they really do vary. Whether it's the fact that the cloud providers have significantly fewer disruptions within their network. Some countries that we've seen have not really been impacted in terms of traffic increases while others have. It really can depend. And so the only way for you to know how your provider is performing or how the key services that you rely on are performing is to have visibility because these days, very often, you don't directly own or manage it. So the only way to ensure that you're getting performance that you need, is to have insight. >> All right, in addition to the report that's coming out, you've got a weekly series I believe that sharing data along with one of our other CUBE Alarms Archana, tell us a little bit about what you're doing there and how that differs. >> Yeah, so we do a weekly podcast. It's just about 15 minutes. It's just to check in to look at what's happened the previous week. So we put this out every Monday and we're looking at whether there have been outages, any interesting news that's taken place, and we'll often go and deep dive on disruptions that have happened. So last week, you probably heard about the CloudFlare outage. It was a pretty big deal. I was getting lots of folks telling me, "Hey, the Internet's down." And it was really just CloudFlare and their DNS service that wasn't available. So we go under the hood and dissect what happened and how it unfolded, and we can show a lot of interesting visualizations around that. >> All right, one last thing, going back to the report, obviously you gather data, you look to be this yearly report, anything along gathering that data, surprises that you've found along this, or putting together the report, are there certain things that longitudinally you might look to do in future studies? >> Yeah, so I do think that maybe it wasn't as surprising to a lot of people, but it was surprising to us given that looking at the same amount of data or same amount of infrastructure that cloud providers were just so dramatically experienced fewer outages. ISP is like 10 times the number of outages as cloud providers. I think going forward, it would be interesting to incorporate more insight into LastMile connectivity, 'cause we're really focusing on backbone networks, really anything other than LastMiles. So, in subsequent reports, we'll fold in some additional insight into LastMile performance as well. >> Excellent. All right, Angelique, I'll let you have the final word, final takeaways you want people to have from this internet performance report. >> You know, I think what you should take away is that if you're able to see how providers are performing, you really can influence how they operate and have a more productive experience working with them. And because these days they're really foundational to most enterprises business, so it's really important to understand the differences between the cloud providers, as well as differences between internet service providers and how that works across different regions. >> All right, well, Angelique, thank you so much for sharing the results of this. Definitely look forward to digging into the data and hearing more from your weekly activities. >> Thanks for having me. >> All right, thank you so much for joining, I'm Stu Miniman and thank you for watching theCUBE. (bright music)