[ 
https://issues.apache.org/jira/browse/CASSANDRA-6477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14625666#comment-14625666
 ] 

Jonathan Ellis commented on CASSANDRA-6477:
-------------------------------------------

Re null handling: as I said initially, just leaving out data for a null value 
is broken.  It's not good for a single column, and it's worse for multiple 
columns.  [It sounds like you are saying it treats (A, null) == (null, B) == 
null which is not what we want either.]

What if we decided that MV partition key will always be encoded as a tuple?  
Does that get around the "partition key cannot be null" problem if we have 
tuple(null) instead?

(I don't see how NOT NULL works except as an extension of [INSERTS 
ONLY|https://issues.apache.org/jira/browse/CASSANDRA-9779], because otherwise 
you either have to introduce possibly racy read-before-write or you have to 
require specifying the NOT NULL column on every UPDATE.)

> Materialized Views (was: Global Indexes)
> ----------------------------------------
>
>                 Key: CASSANDRA-6477
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-6477
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: API, Core
>            Reporter: Jonathan Ellis
>            Assignee: Carl Yeksigian
>              Labels: cql
>             Fix For: 3.0 beta 1
>
>         Attachments: test-view-data.sh, users.yaml
>
>
> Local indexes are suitable for low-cardinality data, where spreading the 
> index across the cluster is a Good Thing.  However, for high-cardinality 
> data, local indexes require querying most nodes in the cluster even if only a 
> handful of rows is returned.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to