Hi Aaron,
I failed to mention that my queries have very high frequency so this is not
a data analytics problem.
That's the reason I wanted to reduce transferring a lot of data to client or
using standard map/reduce solutions that can do the job but add too much
latency.
Pig or hive are appropria
Couple of thoughts...
- Use only the custom secondary indexes you have built using your categories.
Pull back all items that match and then do additional filtering and sorting
client side.
- If you really need very flexible query semantics and the use case allows for
it consider using somethi
Hi,
I would like to combine custom secondary index with Cassandra's
built-in secondary indexes.
Custom secondary index uses multiple columns and is based on
historical but incomplete info about matching data in target column family.
It produces list of "candidate" rows which
could be too large to
