Cases2 cases

1 to 2 of 2

1. IL-0133 A storm shut down cooling at Nebius's us-central1 data center, and a disk hot-plug bug kept Managed Kubernetes control planes down for hours after power returned

   Nebius2026

   Symptoms

   On 19 August 2026 the first equipment-level signs in Nebius's us-central1 region appeared around 09:40 UTC: the InfiniBand fabric degraded and GPU over-temperature alerts began to accumulate. Customer reports followed shortly after 10:00. At 10:15 on-call engineers were paged on critical component temperature alerts, and the incident was declared at 10:21.

   Control plane unresponsiveService outageMonitoring outage

   Trigger

   A storm-related event at the data center building took the building management system offline and shut down the chilled-water cooling loop.

   Root cause

   Dependency failure

   Component

   Data center power and coolingCloud block storageetcdInfiniBand fabric

   Contributing factor

   Single point of failureMonitoring blind spotNo alert on limitNo runbookOverlapping failures

   Detection

   No facility-level alert fired, because the storm also disabled the building management system that monitors the facility. Engineers acted on component over-temperature alerts that paged on-call at 10:15 UTC, after InfiniBand degradation, GPU over-temperature alerts and customer reports; the temperature rise from about 09:15 was visible on dashboards but had no alert.

   Remediation

   • had the facility operator switch cooling to manual bypass mode and opened rack rear doors to vent heat
   • re-energised tripped racks, control-plane racks first
   • restored the regional infrastructure control plane by manual intervention
   • restarted block-storage host components to clear stale disk attachment state
   • raised the VM recovery rate limits by hand for the region-scale recovery
   • started the affected Managed Kubernetes control-plane instances by hand
   • applied a controlled stop and start to control-plane instances whose data disk had been left unattached
   • identified and started 104 customer VMs that the incident had stopped

   Action items

   • Complete a joint resilience review with the facility operator covering redundancy of facility-management and monitoring systems, cooling that keeps running when those fail, electrical protection of cooling equipment and on-site response
   • Add facility-level environmental alerting on cooling plant state and room temperature trend that pages on-call independently of the building management system
   • Set up a documented emergency escalation path with the facility operator, with an on-call contact tree and guaranteed response times
   • Define an emergency heat-load reduction procedure with criteria, ownership and mechanics for shutting down or degrading workloads during a cooling loss
   • Establish a dependency-ordered region cold-start procedure with clear ownership for each service layer
   • Run regular region-scale recovery drills that cover VMs, storage attachments and Managed Kubernetes clusters
   • Fix the disk-attachment bug that extended the Managed Kubernetes outage and deploy the fix to all production regions
   • Improve the recovery policies Compute users can configure, and move all Managed Kubernetes control-plane nodes to the improved policy
   • Automate reconciliation of stale storage attachment state after abrupt host failures
   • Remove in-region dependencies from monitoring startup and recovery, and extend cross-region monitoring
   • Make regional health checks used for global traffic management reflect real service availability

   Lessons

   Facility monitoring that runs on the same system it watches goes silent in exactly the failure it should report, and with no alert on the room temperature trend there was no practical window to shed heat load in a controlled way. A cold start of a whole region exercised recovery paths built for isolated failures: single-attempt VM recovery, conservative rate limits and stale disk attachments all needed manual work. The longest part of the outage was restoring customer resources, Managed Kubernetes control planes above all, so Nebius is extending its recovery drills to that layer.

   View case
2. IL-0350 Hosts in an IONOS TXL cluster lost connectivity when an InfiniBand fabric resweep failed on a full multicast table

   IONOS Cloud2026

   Symptoms

   On 3 March 2026, during a scheduled maintenance to replace a faulty switch in a cluster in IONOS Cloud's TXL data center, monitoring alerts fired and the hosts in that cluster lost network connectivity. Customers saw intermittent network degradation or outages, and the Data Center Designer (DCD) web frontend and the AI Model Hub became unreachable.

   Network unreachableLink flappingService outage

   Trigger

   During a scheduled maintenance to replace a faulty switch, Topology 2 of 2 in the affected cluster was disabled as planned, and a link flap on a switch in Topology 1 then flooded the InfiniBand control plane with alerts and started a fabric resweep.

   Root cause

   Capacity exhaustion

   Component

   InfiniBand fabricNetwork control plane (SDN)BGP routingNIC, cable and switch port linkModel hubDatabase

   Contributing factor

   Automation amplificationExcessive blast radius

   Detection

   IONOS began investigating monitoring alerts in TXL; the first status update was posted at 11:36 UTC.

   Remediation

   • Switched from the affected control plane to its standby
   • Rebooted the locked-up gateways
   • Recovered Topology 1 from the scheduled maintenance
   • Paused provisioning in the cluster that had a backlog of queued jobs
   • Re-enabled provisioning once network connectivity improved and watched job execution
   • Re-established connectivity for the systems that support the DCD

   Action items

   • Adjusted the control-plane configuration to use less aggressive failover timers and reduced heavy sweep activity, lowering the load on the forwarding table
   • Reduced the number of multicast group memberships to relieve the forwarding table
   • Set up continuous monitoring of forwarding table use and control-plane memory with alerts before capacity limits are reached
   • Move the control plane to dedicated high-performance servers, separate from data-plane traffic
   • Audit IPv6 configurations and IPoIB driver settings
   • Modernise the interconnect fabric, including switches, gateways and drivers
   • Hardened IAM maintenance protocols so that critical identity services keep working when the primary Cloud APIs are temporarily unavailable
   • Move IAM to a dedicated platform without a dependency on the standard managed Kubernetes clusters
   • Introduce a dedicated provisioning provider for DBaaS and Managed Kubernetes
   • Add automated circuit breakers to the provisioning engine against overload and queue buildup from self-healing services
   • Improve visibility and control over job execution and prioritisation in the provisioning queue
   • Change the provisioning queue architecture to decouple services and reduce spillover

   Lessons

   IONOS identified the linking of services through the shared provisioning queue as a key reason the degradation lasted as long as it did. Alongside making each service more robust, it is decoupling the provisioning queue and addressing dependencies between services to reduce the blast radius of future incidents.

   View case

By root cause2