The issue has been resolved.
Scheduling delay and api request error rates are back to normal. We are monitoring closely.
We identified a problem with a third party provider and working on mitigating the issue.
We are investigating the issue.