When you create a drive, it is allocated to a randomly chosen physical host that has enough available storage capacity. This ensures even distribution across our infrastructure.
When a server is started, the platform tries to place it on the same physical host as its attached drives. This allows direct, low-latency access to your data. If the same host is not available (for example, due to resource constraints), the server is placed on a different host and accesses its drives over a high-speed peer-to-peer iSCSI connection.
This architecture ensures that your servers and drives are always accessible while optimizing for performance whenever possible.