ML Pod Startup Failures
Mar 30, 14:30 UTCMar 30, 14:30 UTC
Duration
0m
Impact
Informational
Root cause
Not disclosed
Scale AI, 90 days
0 incidents
Affected
Not listed by the vendor.
Final update
There was an outage with a service that manages the metadata for models from 7:06am PT to 7:50am PT. It prevented model services from starting up new pods. ML services that received security updates, deployed a new version, or attempted to scale up were impacted. There was an additional 8 to 12 minutes of service degradation after restoring the impacted ML services to catch up on missed requests.
Timeline
- Resolved · Mar 30, 14:30 UTC
There was an outage with a service that manages the metadata for models from 7:06am PT to 7:50am PT. It prevented model services from starting up new pods. ML services that received security updates, deployed a new version, or attempted to scale up were impacted. There was an additional 8 to 12 minutes of service degradation after restoring the impacted ML services to catch up on missed requests.