Data transform method and data transformer

摘要

A data transform method and a data transformer. The method includes: importing a data transform rule; acquiring from the data transform rule a source data definition, a destination data definition and a data transform rule definition; predicting resource energy consumption parameters of a data transform node server according to the source data definition, the destination data definition and the data transform rule definition; and deploying a resource energy consumption optimization policy of the data transform node server according to the predicted resource energy consumption parameters of the data transform node server.

主权项

1. A data transform method, comprising:
importing a data transform rule; acquiring, directly from the data transform rule, a source data definition, a destination data definition and a data transform rule definition, wherein the data transform rule is used for specifying how data is transformed from a form described by the source data definition to a form described by the destination data definition; predicting values of resource energy consumption parameters of a data transform node server according to the source data definition, the destination data definition and the data transform rule definition, wherein the resource energy consumption parameters include at least one of CPU energy consumption, memory utilization rate, and disk I/O utilization rate, wherein predicting values of resource energy consumption parameters further comprises:
predicting the values of resource energy consumption parameters of the data transform node server in a data extracting phase, a data processing phase and a data storing phase; andidentifying an operating type of the data transform rule according to the data transform rule definition, wherein the operating type comprises at least one of direct copying, data type conversion and macro conversion; predicting the memory utilization rate of the data transform node server in the data processing phase according to the size of data read and written at each time in the data processing phase and the number of intermediate values, wherein the operating type is macro conversion; and deploying a resource energy consumption optimization policy of the data transform node server according to the predicted values of resource energy consumption parameters.