The invention relates to the technical field of data processing, in particular to a data storage method and apparatus for a cross-machine-room-based Hadoop cluster, and the data storage method and apparatus is used for solving the problem of relatively large data transmission flow generated during storage of data in the cross-machine-room-based Hadoop cluster in the prior art. The method comprises the steps of determining a task type used for calculating data in a data table after the data table needed to be written is received; and storing the data meeting a same preset condition or all the data in the data table into a same machine room according to the determined task type. According to the technical scheme, the data meeting the same preset condition or all the data in the data table can be stored in the same machine room, so that the same data is not stored in a plurality of machine rooms and the cross-machine-room data flow generated in the storage process is avoided.