[
https://issues.apache.org/jira/browse/CASSANDRA-6477?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14631440#comment-14631440
]
Sylvain Lebresne commented on CASSANDRA-6477:
---------------------------------------------
I don't really think the cost of replaying the coordinator BL matters that
much. We'll only if less than a quorum of node don't answer a particular query,
which should be pretty rare unless you have bigger problems with your cluster.
And given the local BL has a cost on every write, even if small, I don't think
that from a performance perspective a local BL is a win.
That said, I hadn't seen we'd decided to go with pairing of base replica to MV
replica. Doing so does justify a local BL (another option has always been to
fan out to every MV replica, and since this ticket desperately miss a good
description of what exact algorithm is actually implemented, I wasn't sure
which option we went with).
> 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)
