Hi guys, I have a few questions about the global index feature that's coming in cassandra 3.0 . I’ve had a discussion about this with carlyeks and belliottsmith and I have gone through the discussion in https://issues.apache.org/jira/browse/CASSANDRA-6477. I have some followup questions regarding this feature. Please provide your insights.
- Is the index going to be a separate table behind the scenes? As a user what are the possibilities w.r.t. data in the main table and the index that I have to account for? E.g. I gather that data may be in the index but not main table. This is especially interesting to understand in the presence of failures. - Related to the above point: What about the cases where the consistency level is set to quorum or all? - Can global index be based on multiple columns? For e.g. in primary key I can have a partition key column and a clustering column. Can I do the same in global secondary index? If yes, then is it going to behave the same way as the composite primary key? - Can I create a GSI after creating a table and having some data in it? Cheers, Ajaya
