is Apache Kylin good alternative for SSRS(SQL Server Reporting Services)?

Question

We have framework for analysis data with the help of cube design for OLAP and warehouse that has ETL connections, all of them are in sqlServer structure and SSRS(SQL Server Reporting Services). Some of our reports take a lot of time to come to a conclusion, we decided to migrate to Hadoop ecosystem. is Apache Kylin a good alternative for this framework ?

Answer 1

First of all, I would compare SQL Server Analysis Services with Kylin and not Reporting Services as Reporting Services is closer to front-end system.

Based on my experience, with SQL Server SSAS you can do a lot of things that you can not do with Kylin.

1. Kylin doesn't have MDX Language Support
2. Kylin doesn't have cube Translations (probably because not supp MDX)
3. Kylin doesn't have cube Perspectives (probably because not supp MDX)
4. Kylin doesn't have the complete functionality on the Kylin UI, with that said there are a lot of things that you can not do via UI and you need to add them in the configuration files. (f.e. to add a new SQL Server data source.)

On the other side, Kylin is a distributed system compared with Microsoft SQL Server with Analysis Services, where you can do some finetuning but it will never be a true distributed system.

Answer 2

Based on my limited experience, most traditional OLAP products have limitations on scalability, and could not work together with Hadoop ecosystem very well. Apache Kylin follows the standard OLAP modeling methodology, and provides sub-second performance on Big Data architecture directly. So it's also the only open source OLAP on Hadoop solution. It worths a try.