[
https://issues.apache.org/jira/browse/KYLIN-3044?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kaige Liu updated KYLIN-3044:
-----------------------------
Attachment: KYLIN-3044-sqlserver-as-datasource.patch
Sqoop splits data to a couple of parts and import them parallel. I add a
property kylin.source.jdbc.sqoop-mapper-num to specify how many splits should
be divided. Sqoop would run a mapper for each split.
To make each mapper gets even input, split column is chosen following some
rules:
1. Prefer ClusteredBy column
2. Prefer DistributedBy column
3. Prefer Partition date column
4. Prefer Higher cardinality column
5. Prefer numeric column
6. Pick a column at first glance
Patch updated.
> Support SQL Server as data source
> ---------------------------------
>
> Key: KYLIN-3044
> URL: https://issues.apache.org/jira/browse/KYLIN-3044
> Project: Kylin
> Issue Type: Task
> Reporter: Kaige Liu
> Assignee: Kaige Liu
> Attachments: KYLIN-3044-sqlserver-as-datasource.patch,
> KYLIN-3044-sqlserver-as-datasource.patch
>
>
> [KYLIN-1351|https://issues.apache.org/jira/browse/KYLIN-1351] has added
> Vertica as data source. Base on the work of KYLIN-1351, I'd like to enable
> SQL Server as data source of kylin.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
