13 outages in August 2026: resolution, management and the provider in the middle
Cloudflare logged 13 incidents in eight days, Namecheap lost DNS management for a day, Proofpoint's email domain stopped resolving. What each one means for a zone on two DNS providers.
Published
August 2026 was a busy month for status pages. Between August 7 and August 14, Cloudflare’s status page logged 13 separate incidents, and in the same week two other companies had outages where DNS was either the victim or the cause. None of them was the kind of global outage that makes the evening news. Together they are a good map of what can go wrong around DNS.
Cloudflare: 13 incidents in eight days
The incidents covered R2 object storage, Durable Objects and Workflows, Workers KV, Workers AI, analytics, email security (an issue related to Spamhaus, the only one rated major), Magic Transit, the MCP server portal, and regional network problems in Istanbul, the UK, the eastern US, Mexico and other regions. Twelve of the thirteen carried a minor-severity label.
Authoritative DNS was not among them. The pattern is still worth noting: a provider that runs many products has many ways to have a bad week, and the network incidents are the ones that can reach DNS in some region.
Namecheap, August 13: DNS management down, resolution up
Storms knocked out the cooling at the Phoenix data center that hosts Namecheap’s core infrastructure. Namecheap powered down more than 5,000 servers to protect them and brought services back over about 28 and a half hours. Hosting, email, the dashboards, billing and DNS management were unavailable.
DNS resolution kept working: zones already published kept answering, because the nameservers sat outside the failed data center. What customers could not do for a day was change a record.
Proofpoint, August 14: a domain that stopped resolving
From about 12:26 to 16:08 UTC, hostnames under pphosted.com, the domain Proofpoint uses to route its customers’ email, stopped resolving. Lookups returned NXDOMAIN and some customers reported that the DNS records for their hosted email were gone. Mail to and from thousands of organizations was delayed or bounced. The public reports do not say which DNS provider was involved or why the records disappeared.
How a two-provider zone behaves here
These three cases are three different failures, and a zone on two DNS providers answers each one differently.
A provider’s region or network has problems (Cloudflare). If the problem reaches a provider’s nameservers, resolvers that time out on them retry the other provider’s nameservers from the NS set. Probes from each probe region show the difference: RedundantDNS can alert when a provider is down from some regions only (probe_degraded) or from enough regions to call it down (provider_down).
A provider’s control plane is down, its nameservers are not (Namecheap). This is the case people forget. With a single provider you cannot change a record until its dashboard and API come back. With RedundantDNS the canonical zone is not at the provider: you still edit it, the change applies to the provider that is up, and the one that is down is marked with a sync error and caught up when its API answers again. For those hours the two providers answer differently for the records you changed, which is usually what you want during an incident (the new record is live somewhere) but not always; the dashboard shows exactly which records differ.
The records themselves are wrong or missing (Proofpoint). If a provider loses or corrupts records on its own, the other provider still has them, and verify flags the provider that drifted so you can reconcile it. If the records were deleted by whoever manages the zone, a second provider will delete them too: the zone is the source of truth. Then the journal and the audit log are what get you back: they show the last good version and who changed it.
A second provider is not a cure for every August. It removes one specific single point of failure, the provider, and makes the others easier to see.
Sources
- Shattered.io, Cloudflare logs 13 outages in 8 days as R2 falters, with links to the Cloudflare status page.
- IncidentHub, The August 13, 2026 Namecheap outage.
- StatusGator, Proofpoint outage on August 14, 2026: DNS failure disrupts email worldwide.