incubator-cassandra-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From J T <>
Subject Non-Unique Indexes, How ?
Date Wed, 10 Nov 2010 23:55:04 GMT

I'm trying to work out a way to support a non-unique index.

For example, lets say I have a contact list, where its possible to have
Names that are the same but are for different people and so should have
different contact entries but I'd want to be able to search on their full
name and get a list of potential matches.

In cassandra, as far as i know, column names and row keys need to be unique
- so unless I some how construct a unique form of the full name to use as a
column name or key value I'm left with using the column value (as opposed to
the name) and the indexing facility in 0.7 - but its not clear to me whether
the 0.7 index facility would support non-unique column values this way.


CF: Contacts (with an index on 'fullname')

key : id1 { fullname : "John Brown", address : "London" }
key : id2 { fullname : "John Brown", address : "Paris"    }

Would the 0.7 index on fullname allow me to lookup the 2 entries if I
searched on "John" or "John Brown" ?



View raw message