Cases13 cases

1 to 5 of 13

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-0023 A training run crashed with an NCCL timeout after one InfiniBand port went down once and recovered

   Databricks2026

   Symptoms

   Seven hours into a distributed training run on Databricks AI, the job failed with an NCCL timeout.

   Training job failureNCCL timeoutLink flapping

   Trigger

   A single InfiniBand port used for RDMA NCCL collectives went down once and recovered.

   Root cause

   Dependency failure

   Component

   NCCLInfiniBand fabricPyTorch

   Contributing factor

   Monitoring blind spotUnreviewed defaults

   Detection

   The training job crashed with the NCCL timeout. The continuous health checks, which watched InfiniBand port flapping by count, had not flagged the single flap.

   Remediation

   • Tuned the NCCL_IB_TIMEOUT defaults to be more resilient to short port-down windows
   • Used the port-down signal to crash the job and restart it from a checkpoint instead of waiting for the watchdog
   • Fed the investigation into gpu-monitor, whose RDMA fabric port-down check is thresholded on cumulative downtime rather than flap count

   Action items

   • Tune the NCCL_IB_TIMEOUT defaults to be more resilient
   • Crash and restart the job from a checkpoint on the port-down signal instead of leaving GPUs idle until the watchdog fires
   • Threshold RDMA fabric port-down detection on cumulative downtime, not flap count, in gpu-monitor

   Lessons

   For training impact the signal that matters is cumulative port downtime, not the number of flaps: one long enough flap can crash a multi-day run. NCCL_IB_TIMEOUT, with an effective default of about seven seconds, decides the outcome long before the PyTorch watchdog timeout does. Health checks need to cover the whole node lifecycle: active checks before workloads start, passive checks for silent degradation under load, and periodic multi-node NCCL fabric checks in between.

   View case
3. IL-0134 A power event at Lambda's AUS01 data center rebooted 246 GPU nodes, all spine switches and the VAST storage, failing customer workloads and leaving stale NFS mounts on VMs

   Lambda2026sev1

   Symptoms

   At 21:06 UTC on 14 February 2026 every GPU node in Lambda's AUS01 data center rebooted at the same moment, without warning or plan, affecting multiple customers. The nodes came back online quickly, but the storage they used had disconnected, and customer workloads failed.

   GPU unavailableStorage disconnected or stale mounts

   Trigger

   A power event at the AUS01 data center caused a loss of power.

   Root cause

   Dependency failure

   Component

   Data center power and coolingInfiniBand fabricShared network filesystem

   Contributing factor

   –

   Remediation

   • restored the compute fabric and InfiniBand connectivity, bringing all spine switches and most leaf switches back online
   • restored part of the UPS redundancy
   • powered up all nodes in AUS01, completed physical checks and started the storage back up
   • restored Starbase and the other storage-dependent services in us-west-3
   • told affected customers to reboot their VMs, or the hypervisor if needed, to clear stale or empty NFS mounts

   Lessons

   A power loss that reboots the network spine and the shared storage together with the GPU nodes takes longer to recover than the nodes themselves. The GPU nodes were back quickly, but the fabric was restored first, storage was started only after physical checks, and some VMs were left with stale NFS mounts that customers were told to fix with a VM reboot or, if that was not enough, a hypervisor reboot.

   View case
4. IL-0187 A brief Spring Cloud Config Server downtime failed every client application's liveness probe, because the actuator health check included the config server, and Kubernetes restarted the pods in a kill loop

   Yonghyun Kim2025

   Symptoms

   While the Spring Cloud Config Server was briefly down for an upgrade, every client application began to answer its liveness probe with a failure saying it could not reach the config server.

   Probe failuresContainer crash loopService outage

   Trigger

   A downtime upgrade of the Spring Cloud Config Server made it stop responding for a short time.

   Root cause

   Dependency failure

   Component

   Liveness and readiness probes

   Contributing factor

   Unreviewed defaultsAutomation amplificationExcessive blast radius

   Lessons

   A liveness probe that includes an external dependency turns a short outage of that dependency into a restart loop of every client. Keep external dependencies out of liveness checks and handle temporary problems with readiness probes. Libraries can add health indicators to Spring Boot's actuator health endpoint automatically, so check what the probe path actually reports. The author ran another Kubernetes environment for two years without liveness probes, relying on strict readiness probes and on the application terminating itself on unrecoverable faults.

   View case
5. IL-0162 LKE users could not reach their cluster control planes or the LKE Dashboard because an earlier database update made the internal LKE DNS service hang

   Linode2025

   Symptoms

   On 1 April 2025 users of Linode Kubernetes Engine (LKE) began having trouble connecting to their clusters' control planes, and the LKE Dashboard was affected as well. Services inside the clusters kept working. The first status update was posted at 20:28 UTC; at 21:33 UTC Linode said it was investigating an issue with DNS functionality within LKE.

   DNS resolution failuresControl plane unresponsive

   Trigger

   An earlier update to a database that the LKE DNS system depends on caused the DNS service to hang.

   Root cause

   Dependency failure

   Component

   Authoritative DNS servers and recordsApplication database

   Contributing factor

   –

   Detection

   The first status update, at 20:28 UTC on 1 April, said Linode was investigating and that users might have difficulty connecting to their cluster control plane. At 21:33 UTC Linode said the issue affected DNS functionality within LKE, and at 22:29 UTC that it had identified the underlying cause.

   Remediation

   • restarted the affected LKE DNS services

   Action items

   • Improve how the DNS services recover from transient connectivity failures to their dependencies
   • Improve monitoring to detect similar problems sooner
   • Update the deployment process to better coordinate changes between dependent systems

   Lessons

   A change to a database that the DNS service depends on can make the DNS service hang, which here cut off external access to the clusters and the LKE Dashboard while services inside the clusters kept working. The DNS service should recover by itself from transient failures of its dependencies, and changes to dependent systems need to be coordinated.

   View case

By root cause1