You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@phoenix.apache.org by "James Taylor (JIRA)" <ji...@apache.org> on 2014/03/19 07:49:43 UTC

[jira] [Updated] (PHOENIX-622) Support multi-tenant tables as views

     [ https://issues.apache.org/jira/browse/PHOENIX-622?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

James Taylor updated PHOENIX-622:
---------------------------------

    Summary: Support multi-tenant tables as views  (was: Support secondary indexes on tenant-specific tables)

> Support multi-tenant tables as views
> ------------------------------------
>
>                 Key: PHOENIX-622
>                 URL: https://issues.apache.org/jira/browse/PHOENIX-622
>             Project: Phoenix
>          Issue Type: Task
>    Affects Versions: 3.0.0, 4.0.0, 5.0.0
>            Reporter: elevine
>            Assignee: James Taylor
>            Priority: Blocker
>             Fix For: 3.0.0, 4.0.0, 5.0.0
>
>
> Support secondary indexes on tenant-specific tables by:
> * creating index on tenant-specific table by inserting indexed columns after the tenant_id and type_id columns, but before the entity_id column either:
>   * in the same underlying HBase physical table, but with a different type_id for the index table. We'd need to be careful when querying the data by filtering out index rows, as they would not conform to the data table row key.
>   * have another separate physical HBase table that stores all index rows. In this case, the same type_id could be used for the index and data table. Probably simplest to create this separate physical HBase table upfront, when the base data table is created.
> Probably the latter approach is better/easier. We could also have different table/column family configurations for the index table which is probably good.
> @elevine - what do you think?



--
This message was sent by Atlassian JIRA
(v6.2#6252)