Summary
- OVHcloud reported increased latency and packet loss in its EU-WEST-PAR region.
- The provider attributed the network disruption to an unexpected underlying infrastructure malfunction.
- Managed Kubernetes customers were also affected by the associated network problem.
OVHcloud restored services in its EU-WEST-PAR public-cloud region after an infrastructure malfunction caused increased latency and packet loss across customer networks.
The network incident began at 17:20 UTC on 22 September and ended at 09:30 UTC on 23 September, according to OVHcloud’s final incident record. Public Cloud services hosted in the Paris region experienced increased latency and packet loss, with customers reporting packet drops while accessing services and networks.
OVHcloud attributed the disruption to an unexpected underlying infrastructure malfunction but did not identify the specific failed component.
The same incident affected the company’s Managed Kubernetes Service. OVHcloud said customers experienced network problems that caused nodes in affected clusters to move in and out of a NotReady state.
That secondary effect illustrates how a fault in shared network infrastructure can propagate beyond basic connectivity. Container orchestration platforms depend on reliable communication between control-plane services, nodes, storage, registries, and applications; packet loss can therefore surface as an application-platform problem even when the originating fault sits lower in the infrastructure stack.
OVHcloud’s status information limits the scope for a deeper root-cause assessment. The provider has not said whether the malfunction involved switching, routing, optical transport, power, or another underlying system, and there is no basis to assign the failure to any particular data centre component.
The incident nevertheless provides another reminder that cloud resilience is ultimately dependent on physical and network infrastructure. Regional architectures can abstract servers and storage from customers, but the services still depend on shared facilities, network paths, and control systems whose failure can affect multiple products at once.
OVHcloud’s Paris region was showing normal operation after restoration. Any fuller assessment of the incident will depend on whether the company publishes additional root-cause detail beyond the operational status updates.
