Quick Answer: Cloudflare, which sits in front of roughly a fifth of all internet traffic, logged 13 separate status-page incidents between August 7 and August 14, 2026, touching R2 object storage, Durable Objects, Workers KV, Workers AI, and regional network performance across four continents. None matched the scale of the company’s catastrophic November 18, 2025 global outage, when a corrupted database file crashed Cloudflare’s core proxy service and took down CDN, dashboard access, and security products simultaneously for nearly six hours. The August cluster reopened a familiar question for engineering teams everywhere: what happens when this much of the internet’s infrastructure sits behind one provider, and something breaks there, over and over, in the same week?
Key Takeaways
- Cloudflare logged 13 incidents in 8 days in August 2026 — 12 rated minor, one (an email delivery disruption tied to a Spamhaus blocklisting) rated major — spanning R2, Workers KV, Workers AI, Magic Transit, and network performance in multiple countries.
- An R2 object storage incident on August 7 left at least one customer reporting roughly 67GB of unrestored data days later, even after Cloudflare’s status page marked the incident resolved.
- Cloudflare’s catastrophic November 18, 2025 outage was caused by a routine database permissions change that corrupted a Bot Management feature file, crashing the company’s core proxy service for nearly six hours and affecting nearly its entire product line simultaneously.
- According to W3Techs, Cloudflare served as the reverse proxy for 24.2% of all websites as of late July 2026 — meaning a regional incident in one country can produce failed logins and broken checkouts for thousands of businesses that never directly chose to depend on Cloudflare.
- None of this has slowed Cloudflare’s business: Q2 2026 revenue reached $696.1 million, up 36% year-over-year, with 74% year-over-year growth in paying customers — figures reported before the August incident cluster.
The August 2026 Incident Streak: 13 Outages in 8 Days
Cloudflare’s own public status history shows the run began on August 7 with a write-availability problem affecting a small number of R2 storage buckets in its Eastern North America (ENAM) region, between 14:52 and 17:02 UTC. A second, unrelated network performance issue hit Istanbul the next day after Cloudflare reported losing dark fiber connectivity in the region. Three relatively quiet days followed before the pace accelerated sharply: August 12 brought a major-severity email delivery disruption tied to a Spamhaus blocklisting, and August 13 alone produced four separate incidents — 503 errors on Magic Transit dashboard and API changes, elevated errors on Workers KV requests, intermittent authentication failures on the MCP Server Portal, and errors affecting specific Workers AI models. August 14 closed the window with four more incidents: a Durable Objects and Workflows availability drop, network congestion in the Eastern US, HTTP 5xx errors across Kuwait, Bangkok, Jakarta, and Dammam simultaneously, and a separate network performance issue in Querétaro, Mexico.
Twelve of the thirteen incidents carried a minor-severity label, and most resolved within hours. But the R2 storage incident on August 7 has drawn particular scrutiny: object storage failures carry higher stakes than a dashboard glitch, since they touch actual customer data rather than just service availability. At least one customer reported in Cloudflare’s own community forum that roughly 67GB of data in an R2 bucket had not been restored days after the incident, even though Cloudflare’s public status page marked it resolved. As of this writing, Cloudflare has not published a formal postmortem confirming or denying permanent data loss in that specific case — a gap between “resolved” on a status page and an individual customer’s missing files that’s a recurring friction point for any provider operating at Cloudflare’s scale.
The Real Benchmark: November 2025’s Near-Global Outage
To put the August cluster in perspective, it’s worth revisiting Cloudflare’s actual worst-case scenario from less than a year earlier. On November 18, 2025, a routine database permissions change triggered a query that generated Cloudflare’s Bot Management feature file with duplicated column metadata, doubling the file’s size well past a 200-feature hard limit built into the company’s core proxy service. The oversized file caused the proxy to crash outright, throwing a Rust panic error across the network starting at 11:20 UTC. The affected product list read like Cloudflare’s entire catalog at once: core CDN and security services, Turnstile (which failed to load entirely), Workers KV, the Cloudflare dashboard itself (locking some customers out mid-incident), Cloudflare Access, and Email Security’s IP reputation lookups. Most services were restored by roughly 14:30 UTC, with full resolution by 17:06 UTC — just under six hours from first impact to close.
Against that benchmark, the August 2026 run of mostly-minor, distributed incidents looks comparatively tame. But the two events share a structural root cause worth naming: a network dense enough that a single bad deploy, a permissions change, or a credential rotation error can ripple across a dozen products simultaneously hasn’t gone away just because November’s specific postmortem fixes shipped. A separate March 2025 R2 incident — caused by an engineer omitting a `–env production` flag during a credential rotation, which left production R2 without valid authentication for 67 minutes — shows this isn’t the first time R2’s write path specifically has drawn public scrutiny in under 18 months.
Why Cloudflare’s Scale Makes Every Incident Bigger News
Cloudflare’s footprint is precisely why a run of individually minor incidents becomes a story worth tracking. According to W3Techs, Cloudflare acted as the reverse proxy for 24.2% of all websites as of late July 2026 — an 84.1% share among sites where a reverse proxy provider could be identified at all — and the company has said it handles more than a fifth of global internet request traffic. That concentration means a regional network hiccup in Querétaro or a Workers KV error spike doesn’t stay contained to Cloudflare’s own dashboards; it surfaces as a slow checkout page, a failed login, or a broken API call for thousands of businesses that never directly chose to depend on any single vendor that closely, they simply used a CDN. Centralizing DNS, CDN, WAF, serverless compute, and object storage behind one provider cuts operational overhead for customers who’d otherwise stitch together five separate vendors — but it also means the blast radius of any single incident can span DNS resolution, page delivery, authentication, and application logic all at once, rather than failing in one isolated, easily-diagnosed layer.
How Cloudflare Compares to AWS, Azure, and Google Cloud
Cloudflare isn’t uniquely troubled among major infrastructure providers during this period. AWS suffered a 28-hour outage in its us-east-1 region in 2026, described as the third major us-east-1 failure in recent memory, and outage-tracking analysis has put AWS’s 2023-2025 average uptime at roughly 99.95% with average recovery times near 2.8 hours, against Azure’s slightly higher 99.97% average uptime but longer average recovery of about 4.2 hours (including an October 2025 global Front Door failure). Google Cloud has generally logged fewer major outages than AWS over the same period, even while posting the fastest revenue growth among the big three hyperscalers. Cloudflare doesn’t publish an aggregate uptime percentage the way SLA-driven hyperscalers do, and it discloses far more granular, service-level incidents publicly than competitors typically do — which makes its status page look busier even though the underlying failure rate may not differ meaningfully from its peers. The comparison isn’t apples-to-apples, but the pattern across all four providers points to the same conclusion: single-region and single-vendor dependency remains the biggest structural reliability risk in modern cloud architecture, regardless of whose logo is on the outage.
What This Means If Your Business Runs on Cloudflare
Full multi-CDN and multi-DNS redundancy is expensive and operationally complex enough that most companies, even ones that can afford it, don’t implement it until after a painful outage forces the question — duplicating WAF rules, cache behavior, and TLS configuration across two providers, then building and regularly testing automated failover logic, is a substantial ongoing engineering investment. For services built directly on Cloudflare-specific products like Workers, Durable Objects, or R2, the multi-provider question gets considerably harder, since there’s no drop-in equivalent at another vendor replicating Cloudflare’s specific programming model — a company built around Durable Objects’ actor model can’t simply redirect DNS to a competitor mid-incident. If you depend on R2 specifically for object storage, treating it the same way you’d treat any single-region storage service — with your own independent backup strategy rather than trusting a single provider’s durability guarantee — is a reasonable precaution given R2’s two public incidents in under 18 months.
Related Technology and Infrastructure Guides
Cloudflare’s reliability issues fit a broader pattern of major platform outages throughout 2026. For a comparable story about a different consumer platform’s repeated 2026 outages, see our coverage of Discord’s outage pattern and its own circular-dependency postmortem. And for a look at a major telecom-specific outage rather than a cloud provider, our coverage of the January 2026 Verizon nationwide outage covers a similar concentration-risk story in a different industry.
Frequently Asked Questions
How many Cloudflare outages happened in August 2026?
Cloudflare’s public status page logged 13 separate incidents between August 7 and August 14, 2026, spanning R2 storage, Durable Objects, Workers KV, Workers AI, email security, and regional network performance. Twelve were rated minor severity, and one — an email delivery disruption tied to a Spamhaus listing on August 12 — was rated major.
Was any customer data permanently lost in the August 2026 R2 outage?
Cloudflare’s official status updates describe the R2 ENAM incident as resolved, but at least one customer reported in the company’s community forum that roughly 67GB of data in an R2 bucket had not been restored days after the incident. Cloudflare had not published a formal postmortem confirming or denying permanent data loss for that case as of this writing.
How does the August 2026 incident cluster compare to Cloudflare’s November 2025 outage?
The November 18, 2025 outage was far larger in scope, simultaneously taking down core CDN, Turnstile, Workers KV, the Cloudflare dashboard, and Access for close to six hours, caused by a database permissions change that corrupted a Bot Management feature file. The August 2026 incidents were smaller and more distributed across the week, and none matched the November event’s blast radius.
What percentage of the internet runs through Cloudflare?
According to W3Techs, Cloudflare served as the reverse proxy for 24.2% of all websites as of late July 2026, an 84.1% share among sites where a reverse proxy provider could be identified at all. Cloudflare has said it handles more than a fifth of global internet request traffic.
What caused Cloudflare’s November 2025 global outage?
A routine database permissions change triggered a query that generated a corrupted, oversized Bot Management feature file, which crashed Cloudflare’s core proxy service across the network. The incident lasted just under six hours from first impact to full resolution.
Did the outages affect Cloudflare’s business or stock?
Cloudflare’s Q2 2026 earnings, reported before the August incident window, showed revenue of $696.1 million (up 36% year-over-year) and 74% year-over-year growth in paying customers. Those results predate the August incidents, so any financial impact from this specific run of outages wouldn’t be visible until the company’s next earnings report.
Should my company use a multi-CDN strategy to avoid Cloudflare outages?
It depends on what’s built on Cloudflare specifically. Classic CDN and DNS failover to a second provider is achievable, though operationally costly. Applications built directly on Cloudflare-specific products like Durable Objects or Workflows are much harder to make portable, since no other vendor offers a drop-in equivalent programming model, which is part of why full multi-provider redundancy remains rare even among large Cloudflare customers.
Is Cloudflare less reliable than AWS or Azure?
The comparison isn’t straightforward. AWS has reported average uptime near 99.95% with roughly 2.8-hour average recovery times over 2023-2025, and Azure roughly 99.97% uptime with about 4.2-hour average recovery. Cloudflare doesn’t publish an equivalent aggregate uptime figure and discloses more granular, service-level incidents publicly than its hyperscaler peers, which makes its status page appear busier even though its underlying failure rate may not be meaningfully worse.
