Incidentsby Ibis[UK-SOUTH1] Partial unavailability of the InfiniBand factory
Dec 24, 20:12 UTCDec 24, 20:12 UTC
Duration
0m
Impact
Informational
Root cause
Not disclosed
Nebius, 90 days
22 incidents
Affected
Not listed by the vendor.
Final update
Our team detected an issue with an InfiniBand switch in the fabric, which caused a partial network interruption for a subset of nodes/workloads. The switch was successfully rebooted and connectivity has been restored. The incident is mitigated; we are monitoring the fabric and investigating the root cause to prevent recurrence.
Timeline
- Resolved · Dec 24, 20:12 UTC
Our team detected an issue with an InfiniBand switch in the fabric, which caused a partial network interruption for a subset of nodes/workloads. The switch was successfully rebooted and connectivity has been restored. The incident is mitigated; we are monitoring the fabric and investigating the root cause to prevent recurrence.
More from Nebius
Full historyMetrics unavailable in Management Console for Monitoring in eu-west1lasted 11mProblems with power for several dozens of nodeslasted 14h 45mIssues with preempted Managed Kubernetes nodes in all regionslasted 24hInfiniBand connectivity issues for GPU clusters in eu-west2lasted 6h 28mCompute VM creation and Object storage are partial unavailablelasted 9mNetwork issues in eu-west2lasted 3h 12m