[hibernate-issues] [Hibernate-JIRA] Commented: (HSEARCH-472) Runtime creation of new shards

[Dobes Vandermeer (JIRA)]

    [ http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-472?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=38915#action_38915 ] 

Dobes Vandermeer commented on HSEARCH-472:
------------------------------------------

@Emmanuel,

In our case we have a multi-tenant architecture and we want to have separate indexes for each tenant since there is no case where a search would span multiple tenants and memory usage is linearly related to index size.

The tenants are provisioned by a user when they sign up a new "business" in our system, which is an entity we create and insert into the database.  So, the creation of new shards would happen frequently, but not as frequently as some other operations.  It is also fine for us to manually request a new shard creation, as long as the error messages are clear when we have forgotten to do so and insert something with a reference to a non-existant shard.

I suppose in an ideal world (for our use case) we'd be able to annotate a collection as full-text searchable and hibernate would automatically create a separate lucene index for each instance of that collection, since clearly searching that collection would not search any other collection at the same time.  Then provide an API for doing full-text-search on a persistent (or even not-yet-persistent!) collection.



> Runtime creation of new shards
> ------------------------------
>
>                 Key: HSEARCH-472
>                 URL: http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-472
>             Project: Hibernate Search
>          Issue Type: New Feature
>          Components: query
>    Affects Versions: 3.2.0.Beta1
>            Reporter: Dobes Vandermeer
>            Assignee: Emmanuel Bernard
>            Priority: Minor
>             Fix For: 3.3.0.Beta3
>
>
> Hibernate Search's shard support has added the ability to select a shard based on a parameter of the query; however, you must use a fixed number of shards.  In order to best support our use case originally reported in http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-470 it would be necessary to add support for dynamically added shards at runtime.
> Also related: http://opensource.atlassian.com/projects/hibernate/browse/HSEARCH-251 and http://www.opensubscriber.com/message/hibernate-dev@lists.jboss.org/9770383.html

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://opensource.atlassian.com/projects/hibernate/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira