We have a desire to make data available for query soon after injest but don't want to have particularly small files on HDFS.
Why is it most injesting approaches don't support append but rather create new files? If you want to query data soon after it is injested is it better to have another database in the pipeline prior to HDFS and batch write from that database to HDFS? Sent from my iPhone --------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@hadoop.apache.org For additional commands, e-mail: user-h...@hadoop.apache.org
