View Full Version : Clarification of limitations of Cassandra Input Step

02-07-2013, 05:52 PM
I have a Keyspace with 2 column families, events and details for those events. The details are linked to the events through a column in both column families. I would like to create a sequence of steps that read rows from the events table filtered on a specific event type (easily done with CQL) and then used the eventID column value in the event data to access the rows in the detail table associated with the event.

This seems like a trivial thing to do in code, but to do this with the Cassandra Input Step I need to be able to have the CQL include a value from the rows collected in a previous step. I found a post on the forums that suggested that binding variables to column data is not supported so it seems like there is no way to do this.

I am just trying to understand if that is really the case and what alternatives there are for performing this sort of sequential fetching from 2 related column families.

Thanks in advance,

02-11-2013, 11:13 PM
Hi Joel,

You are correct. Whilst it is possible to use variables in the CQL query specification, the Cassandra input step cannot, at present, accept incoming rows and bind field values to variables. If this functionality is important to you please create an improvement JIRA and vote for it.


03-04-2013, 11:47 AM
I do agree that it is an useful feature to have the values selected from the previous step. I will also vote for it if you create an improvement JIRA :)