This incident has been resolved.
Beginning at approximately 10:10 UTC, the Rownd token API began experiencing increased latency. Authenticated users may have felt impacts due to timeouts when attempting to refresh an access token that had expired.
The root cause was identified and tracked to a non-critical, downstream, third-party component failure.
A fix was implemented to bypass the downstream component in the event of failure, which would resolve the issue and prevent it from recurring.
The fix was deployed at approximately 13:00 UTC and token refresh response times began to improve immediately.
We are monitoring the situation throughout the day to ensure that the API continues to operate as expected.
