The drop was due to maintenance tasks affecting the versions of the service components; it took longer than expected, but it did not affect any customer data, so the disruption was solely related to the service. Everything is now working correctly and is in an operational state.