I'm performing a superSlice operation on a column family that looks like this:
SuperColumn Family: UserActivity{ User1{ Activity1{ //activity detail columns } Activity2{ //activity detail columns } . . . . } }
Test case: 1. Inserted 5 users each with 10 activities in the SCF (super column family). 2. Executed a superSlice query to fetch the top 5 activities for each user.
This works fine in the first attempt. I then deleted all the details.
I immediately re-ran the use case. This time the superSlice returned all 10 activities for each user instead of the 5 as per the slice range specification.
I'm using the Hector API for Cassandra, but I'm seeing the same issue even with a plain thrift call.
Has anyone encountered this issue? Is there any way to get around this?
- Sagar