Today between 22:16 and 23:05 UTC clients calling Microsoft Graph might have experienced increased latency in the North America and APAC geographic regions. This resulted in some calls to Microsoft Graph timing out or returning 503 HTTP responses.
A performance regression was introduced during the rollout of a scheduled change, which affected several regions. Recovery from the incident started as 22:55 UTC and it was fully resolved at 23:05 UTC.
We’re currently investigating to understand the root cause of this incident so that we can ensure that it does not occur again in the future. We apologize for any inconvenience.
The Microsoft Graph team