[
https://issues.apache.org/jira/browse/USERGRID-396?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Todd Nine updated USERGRID-396:
-------------------------------
Description:
Current instance of tools don't work correctly with 2.0.
This should have a RESTful endpoint that is trigged with a PUT. We should have
something like this endpoint.
Check all unique values across the entire system => PUT
/system/database/unqiuecheck
Check all unique values across a single applications => PUT
/system/database/uniquecheck/application/[appid]
Check all unique values in a single collection => PUT
/system/database/uniquecheck/application/[appid]/[collectionname]
The unique check will need to audit the unique index, and ensure that the
unique value exists for the entity specified.
Also, we should add a unique verification read/repair as a separate ticket.
This should hopefully bring unique indexes into a consistent state after
several misses when we have an orphaned unique value.
was:Current instance of tools don't work correctly with 2.0.
> Create UniqueIndexCleanup tool that works with 2.0
> --------------------------------------------------
>
> Key: USERGRID-396
> URL: https://issues.apache.org/jira/browse/USERGRID-396
> Project: Usergrid
> Issue Type: Story
> Reporter: George Reyes
>
> Current instance of tools don't work correctly with 2.0.
> This should have a RESTful endpoint that is trigged with a PUT. We should
> have something like this endpoint.
> Check all unique values across the entire system => PUT
> /system/database/unqiuecheck
> Check all unique values across a single applications => PUT
> /system/database/uniquecheck/application/[appid]
> Check all unique values in a single collection => PUT
> /system/database/uniquecheck/application/[appid]/[collectionname]
> The unique check will need to audit the unique index, and ensure that the
> unique value exists for the entity specified.
> Also, we should add a unique verification read/repair as a separate ticket.
> This should hopefully bring unique indexes into a consistent state after
> several misses when we have an orphaned unique value.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
