Abstract: Systems and methods for providing updates to a data store are described. A data store may receive, process and distribute update information in a substantially orderless manner. This may include implementing an “eventually consistent” design pattern, in which all operations include an object identifier associated with one of the objects in the data store, an information attribute identifier, an initial state, and a final state. Examples may include comparing the initial state included in the update instruction to a current state of one of the objects. If the initial state and the current state are the same, the current state of the object may be modified to match the final state of the update instruction. If the initial state and the current state are not the same, the update instruction may be sent to a redrive queue, from which the update may be reprocessed at a later time.