[ 
https://issues.apache.org/jira/browse/CASSANDRA-6477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14631967#comment-14631967
 ] 

Jonathan Ellis edited comment on CASSANDRA-6477 at 7/17/15 9:46 PM:
--------------------------------------------------------------------

I'm still skeptical that sync MV updates are useful.  I get that people would 
like to have them, but it's a CAP problem.  You only get sync updates if you 
are willing to throw out availability very very quickly.  Async replication to 
MV may be surprising, but UnavailableException for CL.ONE writes when a single 
machine goes down is more surprising.

(FWIW I have consistently told people that our design goal is for MV to be 
eventually consistent with the base data and everyone has been fine with that.)

So: we can keep sync as an option for the future, but for 3.0 we should not 
scope creep this.  We have designed it as async which is the right thing for 
90% of use cases and that is what we should ship.


was (Author: jbellis):
I'm still skeptical that sync MV updates are useful.  I get that people would 
like to have them, but it's a CAP problem.  You only get sync updates if you 
are willing to throw out availability very very quickly.

So: we can keep sync as an option for the future, but for 3.0 we should not 
scope creep this.  We have designed it as async which is the right thing for 
90% of use cases and that is what we should ship.

> 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