Nebius outage history
We hold 55 Nebius incidents from Nebius's status page, the newest from Nov 9, 2026. 55 of the 55 listed carry Nebius's own update timeline. Each one links to a page with the vendor's updates and what our five-minute checks recorded while it was open.
Uptime & history
Tracking since Jun 22, 2026% of our checks that found Nebius fully operational — a degraded check counts against it, because a slow or partly-broken service is not a working one. Down and degraded time are counted at our 5-minute check resolution, so both are approximate. Longer windows fill in as data accrues — we never show uptime we haven't measured.
- Compute Cloud: some hosts unreachable in eu-north1Nov 9, 2026 → Nov 9, 2026 · resolved
- Network problemsNov 6, 2026 → Nov 6, 2026 · resolved
- Token Factory: some models unavailable in eu-north1Nov 5, 2026 → Nov 5, 2026 · resolved
- Public API gateway partially unavailable in US-NORTH1 and US-CENTRAL1Nov 2, 2026 → Nov 2, 2026 · resolved
- Token Factory: some dedicated endpoints partially unavailableOct 30, 2026 → Oct 31, 2026 · resolved
- Compute Cloud: disk images and snapshots partial unavailabilityOct 28, 2026 → Oct 28, 2026 · resolved
- Token Factory 503 errors in eu-west2Oct 24, 2026 → Oct 24, 2026 · resolved
- Partial IAM and console unavailabilityOct 18, 2026 → Oct 18, 2026 · resolved
- CLI login issuesOct 16, 2026 → Oct 16, 2026 · resolved
- InfiniBand connectivity issues for GPU clusters in eu-west2Oct 15, 2026 → Oct 15, 2026 · resolved
- Connectivity issues between eu-south1 and other Nebius regionsOct 14, 2026 → Oct 15, 2026 · resolved
- InfiniBand connectivity issues for GPU clusters in eu-west2Oct 14, 2026 → Oct 14, 2026 · resolved
How Nebius fails
Since Jun 30, 2026, Nebius's status page has marked 22 Nebius components as impaired, across 16,662 of our five-minute readings. Managed Service for Kubernetes® accounts for the most: 12,922 readings (about 1,077 hours) on 9 days, most recently on Oct 7, 2026. Next is Compute Cloud, with 1,009 readings on 17 days. We hold 55 Nebius incidents from Nebius's status page, the earliest from Apr 9, 2026. Of the 48 that Nebius closed with an end time, the median lasted 1 h 36 min.
| Component | Outage | Degraded | Maint. | Days | Last |
|---|---|---|---|---|---|
| Managed Service for Kubernetes® | — | 12922 | 861 | 9 | Oct 7, 2026 |
| Compute Cloud | — | 1009 | 861 | 17 | Nov 9, 2026 |
| Token Factory | 99 | 163 | 861 | 6 | Nov 6, 2026 |
| Object Storage | — | 215 | — | 6 | Oct 14, 2026 |
| API | — | 209 | — | 5 | Nov 2, 2026 |
| Managed Service for SkyPilot | — | 198 | — | 3 | Sep 15, 2026 |
| Management Console | — | 181 | — | 3 | Oct 15, 2026 |
| Managed Service for MLflow | — | 170 | 861 | 2 | Aug 19, 2026 |
Incidents by month: Jun 2026: 8 · Jul 2026: 5 · Aug 2026: 5 · Sep 2026: 9 · Oct 2026: 13 · Nov 2026: 4
Counts are five-minute readings of Nebius's own component statuses, recorded only when a component is not operational. Hours are readings times five minutes, so a gap in our recording would undercount, never overcount.
Every recorded incident
November 2026 · 4
- Nov 9, 2026Compute Cloud: some hosts unreachable in eu-north1major1 h 48 min3 updates
- Nov 6, 2026Network problemscritical8 h 13 min3 updates
- Nov 5, 2026Token Factory: some models unavailable in eu-north1major36 min2 updates
- Nov 2, 2026Public API gateway partially unavailable in US-NORTH1 and US-CENTRAL1minor3 h 6 min3 updates
October 2026 · 13
- Oct 30, 2026Token Factory: some dedicated endpoints partially unavailableminor7 h 17 min3 updates
- Oct 28, 2026Compute Cloud: disk images and snapshots partial unavailabilitymajor6 h 55 min7 updates
- Oct 24, 2026Token Factory 503 errors in eu-west2major1 h 2 min3 updates
- Oct 18, 2026Partial IAM and console unavailabilitynone—1 update
- Oct 16, 2026CLI login issuesmajor11 min2 updates
- Oct 15, 2026InfiniBand connectivity issues for GPU clusters in eu-west2major10 h 38 min4 updates
- Oct 14, 2026Connectivity issues between eu-south1 and other Nebius regionsminor9 h 9 min4 updates
- Oct 14, 2026InfiniBand connectivity issues for GPU clusters in eu-west2major58 min2 updates
- Oct 8, 2026Metrics unavailable in Management Console for Monitoring in eu-west1minor28 min3 updates
- Oct 7, 2026Problems with power for several dozens of nodesmajor14 h 12 min2 updates
- Oct 2, 2026InfiniBand connectivity issues for GPU clusters in eu-west2minor6 h 28 min4 updates
- Oct 2, 2026Issues with preempted Managed Kubernetes nodes in all regionsminor72 h 23 min3 updates
- Oct 1, 2026Compute VM creation and Object storage are partial unavailablemajor33 min3 updates
September 2026 · 9
- Sep 23, 2026Network issues in eu-west2major3 h 12 min5 updates
- Sep 22, 2026Object storage unavailable in eu-west1major—1 update
- Sep 17, 2026Compute services are partially unavailable for customersmajor43 min2 updates
- Sep 17, 2026Resuming IAM accounts is unavailablenone58 min3 updates
- Sep 16, 2026[eu-west1] Metrics for resources in this region are partially unavailablemajor1 h 29 min2 updates
- Sep 16, 2026[us-central1] Object Storage is partially unavailablemajor4 h 57 min3 updates
- Sep 14, 2026Managed Service for SkyPilot: cluster creation and update requests failing in eu-north1major24 h 33 min3 updates
- Sep 9, 2026Unstable compute VM createminor24 min3 updates
- Sep 1, 2026Connectivity issues for newly created worker nodesmajor15 min4 updates
August 2026 · 5
- Aug 19, 2026Region us-central1 partial degradationmajor21 h 33 min13 updates
- Aug 17, 2026Issues logging to the Management Consoleminor—1 update
- Aug 15, 2026Partial InfiniBand unavailabilitymajor3 h 17 min2 updates
- Aug 12, 2026External VPC connectivity issuesmajor26 min3 updates
- Aug 5, 2026Creation of new VMs is brokennone1 h 18 min4 updates
July 2026 · 5
- Jul 27, 2026Network issuesmajor1 h 33 min4 updates
- Jul 22, 2026[eu-north1] Network timeouts affecting multiple servicesmajor—1 update
- Jul 14, 2026InfiniBand switch failure affecting GPU workloads in the eu-north1-c fabric in the eu-north1 regionmajor22 min2 updates
- Jul 3, 2026Managed SkyPilot OIDC Login is unavailableminor—1 update
- Jul 1, 2026Monitoring, Object Storage and Token Factory services are partially unavailablemajor1 h 44 min5 updates
June 2026 · 8
- Jun 30, 2026New virtual machines are not being createdminor4 min3 updates
- Jun 24, 2026Object Storage service partially unavailable in US-CENTRAL1 regionmajor1 h 37 min4 updates
- Jun 16, 2026One InfiniBand switch failureminor32 min2 updates
- Jun 15, 2026Network Connectivity Issuesminor21 min2 updates
- Jun 10, 2026Compute instances startup issuesnone2 h 33 min2 updates
- Jun 9, 2026Some public IPs unreachable in eu-north1minor—1 update
- Jun 4, 2026Network Connectivity Issuesmajor1 h 28 min3 updates
- Jun 2, 2026Unable to allocate new public IPv4 addresses in eu-north1minor24 h 13 min4 updates
May 2026 · 5
- May 29, 2026Power issuesnone—1 update
- May 19, 2026Unable to perform operations on a small subset of virtual machinesminor1 h 55 min3 updates
- May 13, 2026InfiniBand performance degradationminor2 h 20 min2 updates
- May 7, 2026Nebius console is unavailablenone24 min3 updates
- May 6, 2026Managed Kubernetes operations can take longer than expectedmajor19 h 6 min4 updates
April 2026 · 6
- Apr 29, 2026Partial unavailability of external models in Token Factorymajor16 h 29 min2 updates
- Apr 24, 2026Partial degradation of Dedicated Endpoints in Token Factoryminor43 min3 updates
- Apr 20, 2026Metrics for VMs are unavailableminor25 min2 updates
- Apr 20, 2026TokenFactory DataLab unavailabilityminor25 h 49 min4 updates
- Apr 15, 2026TokenFactory — Some Serverless Models Partially Unavailable in us-central1 Regionminor1 h 36 min4 updates
- Apr 9, 2026Degraded logs, metrics and audit trails availabilityminor—1 update
Incidents and their updates are Nebius's own words from its status page, recorded every five minutes; we never add an inferred cause. AIDownCheck is independent and not affiliated with Nebius.