hadoop - 亚马逊流式弹性mapreduce获取输入文件名并设置输出

Question

我正在使用亚马逊流媒体功能，是否可以获得文件输入名称？

我也可以设置输出，所以每个新行都不是一个新文件吗？

Answer 1

You're not using this correctly, if this is your question. Your mapper should not care which file it's working on. It gets a block of data from the entire set matching your --input argument, sends it to the mapper on the STDIN. The mapper should process the input and output a key\t*val1*,val2,... Those are then passed to (one of) your reducer on its STDIN.