Compute, storage and managed Kubernetes in the European provider’s largest region have been unavailable since 06:12 CET. Hospitals, three city councils and hundreds of firms are affected. We are following the recovery and asking what it means for sovereign-cloud promises.
Live · Frankfurt
Lena HartwigIn Frankfurt
Tomasz WronaCustomers desk, WarsawProvider status
Recovery under way. Halden Cloud says object storage is back for 70% of customers and estimates full service by 13:00 CET.
AnalysisSovereign cloud, single region: the resilience gap nobody priced in
ExplainerNIS2 incident reporting: what the 24-hour and 72-hour clocks require
GuideMulti-region on a European provider: a set-up for small teamsChecked against the provider status page at 10:31 CET
Pinned
Key moment
Halden Cloud says cluster APIs in Frankfurt are responding again. Worker nodes are being restarted in batches; customers should expect workloads to reschedule over the next hour.
Lena Hartwig · Source: provider status pageLink to this update
Correction
We earlier said two city councils were affected. A third, in the Rhine-Main area, has since confirmed its permit portal is offline. The entry below has been updated.
Editor’s deskLink to this update
Two smaller European providers have posted offers of free capacity for a week to customers who need to fail over. We have not verified how many customers have taken them up.
Key moment
Buckets are readable again for most customers, according to Halden Cloud. Writes remain disabled in the affected hall while engineers verify replication consistency.
Lena Hartwig · Source: provider statementLink to this update
We asked readers running on Halden Frankfurt how they were affected. Responses came from hospitals, logistics firms, municipal IT teams and software companies.
Illustrative data. Source: EuroTechBrief reader survey of affected customers, n=214, 08:00 to 09:30 CET.
Ana Ribeiro, data deskLink to this update
On a call with customers and press, the company’s CTO explained why the failover stalled: a configuration check blocked the move because the two remaining halls would have run above their safe power limit.
The safety check did what it was designed to do. The design was wrong, and that is on us.
Key moment
Halden Cloud confirmed that a chiller fault in hall B forced an automatic shutdown. Automated failover to halls A and C began but was halted by a capacity safeguard.
Lena Hartwig · Source: provider statementLink to this update
Online permit services in three councils are offline. Two hospital groups say clinical systems are running on local fallbacks; appointment booking and staff email are down.
Halden Cloud is one of the providers most often named in public tenders that require EU ownership. Today tests a claim at the heart of the sovereignty debate: that local control can come with hyperscaler-grade resilience.
Ilse Varga, analysisLink to this update
Key moment
Halden Cloud says it has notified its national authority, as the directive requires within 24 hours of a significant incident. A fuller notification is due within 72 hours.
Ilse Varga · Source: provider statementLink to this update
Customers who also run in Halden’s Oslo or Lisbon regions report that traffic shifted within minutes. Several said they had tested failover in the past quarter.
The provider’s status page moved all Frankfurt services to “major outage” at 06:31, nineteen minutes after the first customer reports.
Customers began reporting failures in Halden Cloud’s Frankfurt region shortly after 06:10 CET. Lena Hartwig is in Frankfurt; Tomasz Wrona is speaking to affected customers.
No key moments yet.
Coverage began at 06:40 CET. Filed under Sovereignty, Self-Hosting, Business.