An approach to power economization in a spoke and hub environment is presented. When a hub receives a data set from a first spoke, the hub writes the data set to a first partition that is associated with the first spoke. After the data transfer is complete, the hub powers down the first partition. If the first spoke sends additional data transfers to the hub while the first partition is powered down, the hub stores the additional data transfers in a second partition. When the first partition is powered up again, the hub moves the data intended for the first partition and that was stored in the second partition. The data is moved from the second partition to the first partition. The hub may monitor the size of the addition data transfers. If the data transfer is large, the hub may power up the first partition, write the data in the first partition, and move to the first partition any data that was intended for the first partition but written to the second partition while the first partition was powered down.