how to create OLAP Cubes on HADOOP data - Impala Database

[#1]

Dear Cloudera Community,

I am looking for advice on how to create  OLAP Cubes on HADOOP data - Impala Database with  Fact and DIMENSIONS tables.

Reference:

https://vision.cloudera.com/building-a-hadoop-data-warehouse-hadoop-101-for-enterprise-data-professi...

Question number 10

Regards

Anis

[#2 · Accepted solution]

Depending on your budget and feelings on opensource there are two options to push the accelerator. I haven't dug into either one to talk to how they do it.

If you wanted to just roll your own I'd imagine it would be following the star schema methodology the Kimball is known for in Impala and then just leveraging Impala/HDFS caching to make it as fast as possible.

http://kylin.apache.org/

or

http://www.atscale.com/

[#3 · Accepted solution]

Depending on your budget and feelings on opensource there are two options to push the accelerator. I haven't dug into either one to talk to how they do it.

If you wanted to just roll your own I'd imagine it would be following the star schema methodology the Kimball is known for in Impala and then just leveraging Impala/HDFS caching to make it as fast as possible.

http://kylin.apache.org/

or

http://www.atscale.com/

[#4 · Accepted solution]

Hi @AnisurRehman,

Did you check out the follow up posts from the series including:

• Ralph Kimball and Kaiser Permanente: Q&A Part I – Hadoop and the Data Warehouse
• Ralph Kimball and Kaiser Permanente: Q&A Part II – Building the Landing Zone