Wondering if it's sane to change the dataType in a processor in chukwa? I have a single log file (web log) that contains various lines....would like to shard the data based on the request URI in the log.
e.g. /chukwa/repos/[clusterName]/dataType1/20100727 /chukwa/repos/[clusterName]/dataType2/20100727 Is this asking for trouble somehow? Using the UTF8 log adaptor.
