[druid-user] Keeping access to old data with less granularity

Hey Victor,

Another option is to index the data twice upfront into two datasources, just with different names and different granularities. You can set a 2 week retention on the high granularity datasource and longer retention on the other datasource.