On July 17th, at approximately 8:50 UTC, we became aware of service degradation affecting a portion of Block Storage volumes within our Newark data center. Our Administrators were able to identify a system within the affected Block Storage cluster which caused cascading I/O delays for a section of this cluster. This was escalated to our Block Storage team, who removed the faulty system from the cluster at which point all I/O returned to a normal running state.
A pending investigation into the specific hardware failure and the associated redundant systems will be completed by our Operations team. No further impact to service will occur as a result of this investigation.