Re: [druid-user] Unable to query

What does the EXPLAIN plan say for that query? What is the output of SELECT * from filterDataStream? Does that column return then?

What about the other question Rachel asked? Does
SELECT * FROM filterDataStream
return data, and show data for rtcBusTypeList?
Also, what is the output of
SELECT COLUMN_NAME FROM INFORMATION_SCHEMA.COLUMNS WHERE TABLE_NAME = ‘filterDataStream’ ?

Do you have any data in that column yet? If not, can you insert a row that has a dummy value? (Also, is this a real-time datasource?)

“Do you have any data in that column yet? If not, can you insert a row that has a dummy value? (Also, is this a real-time datasource?)”

I ask because for other datasource types, druid will inspect the segments to find the schema. For realtime, it should consult the realtime spec as well.

How is the data ingested? If it’s not realtime (streaming), you might need to add dummy values in general, or for each day or segment granularity. Druid examines the segments for the query to get the schema, so it could be only looking in the interval you specified - I’m not sure.

This may be interesting to read:

I suspect that your first SELECT is covering segments (period of time) that doesn’t have that column in the source data (you have the last eight days) whereas the long JSON version covers all time.