You could adapt the log-synth program to produce files in parquet. At the same time, I would recommend that you build a map-reduce version as well so that you can generate much bigger files.
On Thu, Dec 19, 2013 at 10:55 PM, YouPeng Yang <[email protected]>wrote: > Hi > I have complied the drill M1 with hadoop-2.2. And I have tested the > sample-data/region.parquet in HDFS.It goes well. > > I want to make a deep test on the performance of dirll. > First thing it need to create a huge size of parquet file. Is there any > tools to convert my flat file or other format file to parquet file? > > > > Regards. >
