Refer to this http://dev.bizo.com/2011/02/columns-in-hive.html
<http://dev.bizo.com/2011/02/columns-in-hive.html>HTH - Youngwoo 2011/3/2 Sunderlin, Mark <mark.sunder...@teamaol.com> > Let us say my log data that I want to place a log file into hive. And > the log file itself looks something like this: > > > > Event_time, event_type, event_data_blob > > > > And the blob data looks like > > “Key1=value1;key2=value2;key3=value3 … keyn=valuen” > > > > This looks like maybe I start like this: > > > > Create table my_log( > > Event_time STRING, > > Event_type INT > > Event_blob MAP<string,string> > > > > How am I doing so far? > > > > Then, I know I need to have some reformatting some on the blob to get it > into the ‘right’ format, as I doubt the equals sign ‘=’ works as a key to > key value map in Hive, and I suspect the semi-colon ‘;’ isn’t what > hive/hadoop wants for a key, value pair separator either. > > > > What I don’t know is what format to place the blob data into so I can then > load it into a hive table. > > > > Anyone got some pointers? > > > > --- > > *Mark E. Sunderlin* > > *Solutions Architect **|AOL **Data Warehouse* > > P: 703-256-6935 | C: 540-327-6222 > > AIM: MESunderlin > > 22000 AOL Way | Dulles, VA | 20166 > > [image: aol-comp-logo] > > >
<<image001.gif>>
