We finished the configuration changes of the Premium.CDN cluster. All system are fully operational.
During the planned maintenance we found out that some nodes of the cluster were on extremely high memory usage. In order to avoid eventual memory issue, we proactively restarted some of the nodes in Europe and North America. We tried to minimize any impact, but short service interruption could be seen around the provided time.
We are going to extend the Premium.CDN cluster configuration. We will add spare nodes on each machine. This operation is not expecting any downtime.