Delayed Replica Training Time

Incident Report for Tavus

Resolved

This incident has been resolved.
Posted Sep 27, 2024 - 20:36 UTC

Monitoring

A fix has been deployed and new replicas are now processing as expected within SLA. Replicas previously impacted may still be completing their training. Our team is continuing to monitor training times.
Posted Sep 27, 2024 - 03:51 UTC

Investigating

Replica training is taking longer than expected due to a GPU outage.
Posted Sep 26, 2024 - 16:27 UTC
This incident affected: Replica Training Service.