Hi,

I would use Kylin as MOLAP and Spark as ROLAP. Also I would like to store the 
aggregates on HBase by using my ETL based on Spark, for this purpose I would 
like to use the metadata API of Kylin and preaggregate data on spark before 
loading. 

In this way I should obtain streaming/realtime analytics while loading.

Is it feasible ?

Regards, 

-- gas


Gaspare Maria

<div>-------- Messaggio originale --------</div><div>Da: "Adunuthula, Seshu" 
<[email protected]> </div><div>Data:05/06/2015  19:50  (GMT+01:00) 
</div><div>A: [email protected] </div><div>Oggetto: Re: Kylin 
using Hbase for Tableau </div><div>
</div>Vineet,

Kylin is a MOLAP implementation on Hbase. Its takes raw data from Hive and
stores the aggregates into Hbase.


Your use case appears to be relational access on top of Hbase. There are a
variety of solutions out there,

https://www.simba.com/connectors/apache-hbase-odbc#overview_content

http://www.splicemachine.com/product/features/


Regards
Seshu Adunuthula




On 6/4/15, 10:39 PM, "Vineet Mishra" <[email protected]> wrote:

>Hi,
>
>I have got a usecase where I want to query some of my aggregated data and
>visualize it on Tableau. I am persisting the data in Hbase and making a
>Simba ODBC connection to Hbase through Tableau.
>
>I was keen to know how Kylin fit in my use case as I could see it even
>provides a ODBC connector which may be a good fit for my usecase, moreover
>I am planning to keep my key as Composite key on which I can query as
>Hbase
>power could be leverage on Key level query.
>
>Looking forward for expert advise.
>
>Thanks in Advance!

Reply via email to