cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stu Hood (JIRA)" <>
Subject [jira] Commented: (CASSANDRA-749) Secondary indices for column families
Date Wed, 24 Mar 2010 06:03:27 GMT


Stu Hood commented on CASSANDRA-749:

Cliff mentioned in #cassandra that he preferred the local secondary index approach. I think
this was because adding more replicas is supposed to be a solution for scaling out reads,
but I don't think it is that simple here.

Rather than utilizing IO efficiently by partitioning, all of those nodes will have approximately
the same stuff cached in memory, because they won't have been assigned anything specific to
hold onto. Also, scaling up to more replicas (than you need for HA) across your entire cluster
in order to improve your average latency seems like a waste of disks/machines.

Additionally, you have to keep all of those replicas consistent, which means every node still
has to answer reads eventually for read repair. Adding nodes doesn't actually eliminate any
queries: it might improve your average latency, but it won't affect your throughput.

> Secondary indices for column families
> -------------------------------------
>                 Key: CASSANDRA-749
>                 URL:
>             Project: Cassandra
>          Issue Type: New Feature
>          Components: Core
>            Reporter: Gary Dusbabek
>            Assignee: Gary Dusbabek
>            Priority: Minor
>             Fix For: 0.8
>         Attachments: 0001-simple-secondary-indices.patch, views-discussion-2.txt, views-discussion.txt

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message