You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Vladimir Ozerov (JIRA)" <ji...@apache.org> on 2016/12/29 09:51:58 UTC
[jira] [Updated] (IGNITE-4510) SQL: co-located query may calculate
target partition in advance in some cases
[ https://issues.apache.org/jira/browse/IGNITE-4510?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Ozerov updated IGNITE-4510:
------------------------------------
Labels: performance sql (was: sql)
> SQL: co-located query may calculate target partition in advance in some cases
> -----------------------------------------------------------------------------
>
> Key: IGNITE-4510
> URL: https://issues.apache.org/jira/browse/IGNITE-4510
> Project: Ignite
> Issue Type: Task
> Components: SQL
> Affects Versions: 1.8
> Reporter: Vladimir Ozerov
> Labels: performance, sql
> Fix For: 2.0
>
>
> Consider the following SQL query with {{distributedJoins=false}}:
> {code}
> SELECT ...
> FROM Employee e
> INNER JOIN Department d on d.id = e.dept_id
> WHERE e.dept_id = ?
> {code}
> If {{dept_id}} is affinity key, and this should be certainly so in case of non-colocated joins (otherwise query will return incorrect result), then we can do the following:
> 1) Determine partition for this affinity key.
> 2) Send query execution request to this partition only.
> Same technique can be applied to {{WHERE e.dept_id IN (?, ...)}} case.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)