Internet infrastructure giant Cloudflare has confirmed that a service interruption on the morning of December 5, 2025, briefly disrupted access to major global sites, including Zoom and LinkedIn. The company says service has now been restored, and the outage was not caused by a cyber attack.
According to Cloudflare, the problem stemmed from a change to how its firewall processes incoming requests. That alteration made the network unavailable for several minutes, affecting a wide range of websites and services globally.
Cloudflare also noted that its internal dashboard and related APIs, which allow software systems to communicate, experienced problems, suggesting the disruption went beyond just customer-facing websites.
This outage marks the second major glitch affecting Cloudflare in less than three weeks, raising fresh concerns about the reliability of cloud infrastructure and the risks of over-dependence on a few backbone providers.
Table of Contents
What went wrong
Cloudflare says the outage resulted from a database change it made during what was meant to be routine maintenance. That change inadvertently caused overload on the company’s systems, which triggered the network-wide disruption.
Specifically, the tweak affected how the firewall handles requests. With the new configuration, incoming traffic became unprocessable for a few critical minutes. Though Cloudflare moved quickly to resolve the issue, the interruption was enough to knock multiple high-profile services offline.
Experts say it may take days or even weeks of investigation before the full root cause can be established; the disruption may stem from a chain reaction set off by the maintenance change rather than a single fault.
Impact across the web
As soon as the outage began, users reported being unable to access sites like LinkedIn and Zoom. In addition, some administrators and developers noted that internal dashboards and APIs they rely on were also unreachable, suggesting the effect went beyond consumer-facing services.
The disruption was brief, described by Cloudflare as lasting only a few minutes, but long enough to create a visible ripple across the internet. The recurrence so soon after a previous incident in November has sparked renewed debate over how much of the web depends on a handful of major infrastructure providers.
Some observers pointed to the incident as emblematic of a deeper structural problem: when too much of internet traffic depends on a single provider, a small mistake can cascade into widespread outages.
Growing concerns about cloud centralisation
Industry watchers say today’s outage is not just an isolated misstep, but part of a growing pattern. In recent weeks and months, other major cloud providers have also experienced disruptions. For instance, earlier in November, a Cloudflare outage affected platforms such as ChatGPT as well as popular online games.
The repeated disruptions are raising questions about whether it is wise for businesses, governments, and individuals to place so much trust in a limited number of cloud infrastructure firms. Each time a large provider experiences downtime, the impact touches not just consumer apps but also critical services like communications, remote work, e-commerce, and even internal enterprise tools.
Some experts suggest the industry must rethink assumptions about “always-on” cloud resilience. As one security analyst put it, the problem is that modern digital architecture often funnels enormous traffic through narrow chokepoints — so when something goes wrong, the effects are huge and widespread.
Cloudflare has apologised for the outage and pledged to review internal processes so as to avoid similar failures in the future. But the recurring incidents have already sparked sharper scrutiny over the concentration of power among a handful of internet gatekeepers.
Join Our Social Media Channels:
WhatsApp: NaijaEyes
Facebook: NaijaEyes
Twitter: NaijaEyes
Instagram: NaijaEyes
TikTok: NaijaEyes
