This is the Internet Report: Pulse Update, where we analyze outages and trends across the Internet through the lens of ThousandEyes Internet and Cloud Intelligence. I’ll be here every other week, sharing the latest outage numbers and highlighting a few interesting outages.
This week, we’re exploring key Internet outage trends and sharing tips for proactively mitigating the impacts of future outages as we head into 2024. As always, you can read the full analysis below or tune in to the podcast for first-hand commentary.
Internet Outages & Trends
2023 saw increases in cloud service provider (CSP) outages, application outages, and the percentage of U.S.-centric outages—all of which point to an evolution in the way outages happen and the need for different strategies to minimize the impact of disruptions.
Read on to learn more about these trends, or use the links below to jump to the sections that most interest you. To learn how ThousandEyes gathers the data referenced in this blog, see the Internet Outages Map.
During 2023, ThousandEyes observed an increase in cloud service provider (CSP) outages, perhaps due to the rise in cloud infrastructure deployed worldwide.
While it’s true that Internet service provider (ISP) outages still make up the majority of outages, according to ThousandEyes’ analysis, CSP outages still represent the second most common type of disruption. Interestingly, the average number of CSP outages that we observed in a given month rose during 2023, while there was a downward trend in the average number of monthly ISP outages. Additionally, over the past year, ThousandEyes has observed a slight shift in the ratio of ISP to CSP outages, changing from 89:11 in 2022 to 83:17 in 2023.
What might be causing CSP outages to trend upwards? The rise in cloud infrastructure deployed worldwide likely plays a role. With the increasing demand for compute and data transmission capacity due to the AI boom, cloud infrastructure has grown exponentially, providing organizations with more options for connecting to the Internet than ever before. To deliver good digital experiences for all your users around the world, it’s important to understand this complex ecosystem and to pay attention to the health of the cloud services you rely on.
In 2023, ThousandEyes also saw an increase in the percentage of observed outages that were U.S.-centric, with U.S.-centric outages reaching 37% compared to 34% in 2022. Additionally, in 2023, the percentage of U.S.-centric outages exceeded 40% for six months out of the year; during 2022, the percentage only reached similar heights for two months (see the charts below).
This phenomenon may be explained by another trend we observed: a rise in smaller, more contained outages.
While the past few years brought many massive outages that had cascading effects on other high-profile apps or services, as 2023 progressed, these outages all but disappeared. Instead, shorter, more contained outages and disruptions proliferated. While previously an outage that started in the U.S. may have spread globally, the impacts now often remain more localized.
In the past, the elements of the Internet ecosystem were connected in such a way that an issue in one part of the network would often have a domino effect on others downstream, resulting in widespread impacts around the globe. As the Internet becomes increasingly resilient, there are now more possible paths available to route around issues.
This shift means that organizations like application owners need to alter their incident response strategies a bit. These shorter, more localized outages have different impacts and require different responses than a large-scale global outage would. It’s important to have a deep understanding of your whole service delivery chain so that you can quickly troubleshoot these issues that come up, identifying whether they’re caused by a more global outage that’s happening or whether the problem is internal. With this knowledge, you can efficiently resolve disruptions and provide your users with a smooth digital experience.
Over the last year, ThousandEyes’ analysis shows that application outages also increased both in frequency and number. While they’re still not as prevalent as network outages, application-related disruptions often appear to have a greater impact. Applications rely on a complex network of interconnected dependencies to function properly, as users expect. As a result, there’s a higher risk for extensive disruption and user impact if any of these service or aggregation points experiences a failure or degradation.
To guard against these little issues that can have big impacts, it's important to understand not only how the application interacts with the network, but also how the network interacts with the application. Take time to understand your end-to-end service delivery chain, identify common aggregation points, and develop a plan to optimize performance and address any potential issues.
Looking Ahead to 2024
As we enter 2024, we are facing a different outage landscape than we had at the beginning of 2023. It is crucial now more than ever to take measures to consistently enhance performance and prepare for potential problems so that you can provide your users with a superior experience.