The status of various Savio components can be found at our reporting dashboard, which shows Savio uptime for the past 24 hours.
View the Savio Status Dashboard
We encourage users to subscribe to alerts on the Dashboard's home page.
Savio System Status Definitions¶
The below table describes essential definitions of various terms used in the dashboard and in system status announcements.
| Component | Description |
|---|---|
| Home Directory Storage | /global/home |
| Scratch Storage | /global/scratch |
| Login Nodes | Shared nodes where users access Savio to compile code, submit jobs, and manage files |
| Compute Nodes | Nodes allocated to run user batch jobs |
| Data Transfer Nodes (DTNs) | Dedicated nodes for moving large datasets to and from Savio |
| Job Scheduler | Slurm |
| Network | Connectivity between Savio, the internet, and other systems |
| Open OnDemand (OOD) | Browser-based access to Savio, including a file browser, terminal, and interactive apps |
| MyBRC User Portal | User portal for managing allocations, accounts, and project membership |
Component Statuses¶
Operational¶
Users can perform all expected functions with the component.
Degraded Performance, Partial Outage, and Major Outage¶
Degraded Performance — Users can perform expected functions but some aspect is impaired: reduced throughput, elevated latency, or intermittent errors that slow work without blocking it. A component may also be marked Degraded when a dependency failure meaningfully limits what users can do with it, even if the component itself is functioning normally.
Partial Outage — Some users are unable to complete expected tasks, or a specific function is unavailable, but the component is not entirely down.
Major Outage — Users are unable to perform expected functions with the component.
| Component | Degraded Performance | Partial Outage | Major Outage |
|---|---|---|---|
| Home Directory Storage | Accessible but with noticeably reduced performance | Not mounted on some login nodes | Inaccessible |
| Scratch Storage | Accessible but with noticeably reduced performance | Some storage targets unavailable; files on those targets are inaccessible | Inaccessible |
| Login Nodes | Some nodes inaccessible or slow, but users can still log in and work normally | Failures widespread enough that users have difficulty accessing Savio | Users cannot log in |
| Compute Nodes | Some nodes unavailable due to hardware failures, but no partition is fully impacted | One or more partitions unavailable | Jobs not running due to a widespread failure |
| Data Transfer Nodes | Transfers working but slower than normal | Some DTNs unavailable, reducing transfer capacity | Cannot transfer data to or from Savio |
| Job Scheduler | Available but slow or intermittently failing | — | Cannot submit or manage batch jobs |
| Network | Elevated latency or packet loss affecting connectivity | Some users cannot reach Savio, or the cluster's internal network is impaired | Savio is unreachable from the internet |
| Open OnDemand (OOD) | Accessible but slow | — | Inaccessible |
| MyBRC User Portal | Accessible but slow | — | Inaccessible |
Cells marked — indicate that the status level is not currently in use for that component.
Under Maintenance¶
A scheduled maintenance window is active. This status is not used for unplanned outages.
Incident Statuses¶
Investigating¶
We are aware of a potential issue and are actively investigating. Scope and severity have not yet been determined.
Identified¶
The cause has been identified and work is underway to resolve it.
Monitoring¶
A fix has been applied. We are monitoring to confirm the issue is fully resolved.
Resolved¶
The issue is resolved and all affected systems are operating normally.