hive-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From J├Ârn Franke <jornfra...@gmail.com>
Subject Re: Indexing
Date Sun, 31 Dec 2017 07:00:55 GMT
Hallo,

It always depends on your use case and you always should do performance tests to verify it
fits your use cases. Hence, I doubt that you find a generic statement on the Hive site. Although
most of the times the internal index of orc will have more advantages, such as less space
usage.

Furthermore, orc (or parquet) require that the data is sorted on the filtering column. 

Hive provides also other relevant features, such as partitioning.

Best regards

> On 31. Dec 2017, at 04:28, Sachit Murarka <connectsachit@gmail.com> wrote:
> 
> 
> Hello,
> I have seen some blog saying that Indexing is not recommended , instead we can use ORC
format. Can you please provide suggestion?
> I could not see any official declaration.
> 
> Kind Regards,
> Sachit Murarka
> 

Mime
View raw message