> On May 31, 2013, 8:30 p.m., Abraham Elmahrek wrote:
> > Seems good except for a few comments and nit picks.
> > 
> > The datetime split logic is when certain intervals are "hotter" than 
> > others. IE: 1000000 rows out of 20000000 exist between the date range of 
> > december 1st to december 31st, but a user is importing the entire years 
> > data, with 12 node. Basically, 1 machine will extract 1000000 rows, while 
> > the others will extract 1000000/11 rows. In the future we could probably 
> > add some upfront analysis of the data to improve distribution.

Sorry... my wording above wasn't very concise... The datetime split logic 
doesn't seem to handle "hot" partitions is what I meant.


- Abraham


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/11537/#review21237
-----------------------------------------------------------


On May 30, 2013, 6:25 p.m., Venkat Ranganathan wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/11537/
> -----------------------------------------------------------
> 
> (Updated May 30, 2013, 6:25 p.m.)
> 
> 
> Review request for Sqoop.
> 
> 
> Description
> -------
> 
> This addresses Boolean, date, time, and timestamp splitters.
> 
> THis also disallows char type splitters as discussed in SQOOP-976
> 
> 
> Diffs
> -----
> 
>   
> connector/connector-generic-jdbc/src/main/java/org/apache/sqoop/connector/jdbc/GenericJdbcImportPartitioner.java
>  f80f30d 
>   
> connector/connector-generic-jdbc/src/test/java/org/apache/sqoop/connector/jdbc/TestImportPartitioner.java
>  ee314d0 
> 
> Diff: https://reviews.apache.org/r/11537/diff/
> 
> 
> Testing
> -------
> 
> Introduced new unit tests to test new functionality
> All tests pass
> 
> 
> Thanks,
> 
> Venkat Ranganathan
> 
>

Reply via email to