Multipleinputs.addinputpath, pay attention.

Multipleinputs.addinputpath Note Blog Category: Hadoop hadoop&http://www.aliyun.com/zixun/aggregation/37954.html "> nbsp; Original, reprint please specify the Source:





use Multipleinputs.addinputpath to add multiple input sources (more than two):


multipleinputs.addinputpath (conf, new Path (Otheargs[0)), Textinputformat.class,joinnodemapper2.class); Multipleinputs.addinputpath (conf, new Path (otheargs[1)), Textinputformat.class,joinmemmapper2.class); Multipleinputs.addinputpath (conf, new Path (otheargs[2)), Textinputformat.class,joincpumapper2.class);


my file format is:


Otheargs[0] point to file format:


key1 Keyname1key2 keyname2


Otheargs[1] point to file format:


key1 Vala1key2 valA2


otheargs[2] point to file format:


key1 Valb1key2 valB2




When the
runs the specified parameters in Elipse, if OTHEARGS[2] is the same as otheargs[1, then multipleinputs only executes one of the addinputpath statements, and the result of my test environment is the execution


Multipleinputs.addinputpath (conf, new Path (otheargs[2)), Textinputformat.class,joincpumapper2.class);


1. Request otheargs[2] and otheargs[1] either.


2.reduce After the build file is:


key1 keyname1 vala1key2 keyname2 vala2key1 keyname1 valb1key2 keyname2 valB2


instead of:


key1 keyname1 valA1 valb1key2 keyname2 valA2 valB2


so to achieve the result of the second join, I 211.html "> feel that I still want to use the MapReduce job flow control."