| 摘要 |
Provided are a data processing method and a central node. The method comprises: according to a first looping function which is provided by a user and is written using a MapReduce computing framework, generating, by a central node, a second looping function, a computation enabling function and a second copying function, wherein the second looping function is used for iteratively invoking a first copying function to copy a plurality of data records in a computing node which need to be processed by a GPU from a memory of the computing node to a video memory of the GPU; a Map computing function in the computation enabling function is used for instructing the GPU to process the data records which the GPU is responsible for processing; and the second copying function is used for copying results of computing the plurality of data records by the GPU from the video memory of the GPU to the memory of the computing node, and therefore, a code which is suitable to operate in a CPU is automatically generated into a code which is suitable to operate in the GPU is achieved, so that a Hadoop programming framework is suitable for conducting data processing in a mixed cluster system. |
