Observed a issue where a service health indicator is red but when you drilling down into its topology, all entities are green. The indicator eventually updated to green but takes more then 20 minutes. The service health indicator is used in dashboards and the delayed in status update results in inaccurate dashboards.
It seems like some time this type of issue can occur where the Service health indicator status does not get updated in timely manner. This issue is fix in the latest release.
The workaround for older releases is to update the service details. In this case service description can be updated which will trigger Service health indicator status update and will correct this issue.