Live / Nebius / Incident
Nebius: InfiniBand connectivity issues for GPU clusters in eu-west2
Nebius · nebius.comMajor incidentResolved2026-10-15 16:12 UTC
Public — what happened
Summary
Nebius reported a major problem with InfiniBand networking links for GPU clusters in its eu-west2 region, with investigation beginning at 05:34 UTC on October 15, 2026. By 09:32 UTC the company said it had found the underlying cause and applied fixes, though it did not describe what that cause was. A later update at 13:13 UTC said the affected systems had been stabilized but recovery was still underway and work on the root cause continued. Nebius marked the incident resolved at 16:12 the same day.
Affected
Compute Cloud
Timeline
- Started — 2026-10-15 05:34 UTC
- Investigating — 2026-10-15 05:34 UTC (2 updates)
- Monitoring — 2026-10-15 13:13 UTC
- Resolved — 2026-10-15 16:12 UTC · 10h 38m after it started
With SquidGrid — where it starts on the route
Independent path health
The status page is the symptom. SquidGrid measures the paths that reach Nebius against their own baselines — whether the routes carrying your traffic degraded, at which handoff, and who owns that stretch.
🔒 Sign in to seeCustomers get the live Provider Assessment Report for Nebius: its full incident track record, what tends to break, and recent public reports of problems.
Included with any paid plan.
Open the Nebius report →
More from Nebius
Covered in The Squid’s Ink, Week 43, 2026, our weekly internet health report.