The incident has now been resolved. The timeline for the elevated errors was 6:22 PT / 13:22 UTC -> 6:39 PT / 13:39 UTC.
We are currently investigating this issue.