Integration of custom service

[IoT-Guy]

Hello everyone,

I hope you can help to get a better understanding, how our solution could work.

We're would like to store a huge amount structured and unstructured data and make it accessible to MS PowerBI. Since we're starting from scratch I would like to consider also the requirements of MS PowerBI for this solution.

The data will be stored on AWS but we're not limited to the AWS services. Furthermore the structured and unstructured data is time-series data and it can be stored separatedly, if we can represented within one database model.

Since I found no information "How to integrate custom services or databases into MS PowerBI?" I not sure, what MS PowerBI expects. Are we fine with a hadoop cluster? Do we necesserily need a REST API etc.?

Can I implemented something like customer connectors in MS PowerBI to integrate my custom service?

What's the best way to make huge amount of data accessible for MS PowerBI also in terms of scaling?

I am properly not the first who wants to integrate a custom service into MS PowerBI, therefore it would be really great, if you could share some of your expierences.

Thanks in advance.

IoT-Guy

[Anonymous · Accepted solution]

@IoT-Guy,

Firstly, which backend database do you use for the AWS service? In Power BI Desktop, you can import data from different kinds of database servers(SQL Server , Oracle, MySQL, DB2, etc.). And you are able to connect to multiple data sources in a single PBIX file. You don't have to use Rest API in this case.

Secondly, it is possible to create an ODBC datas source for hadoop cluster, and then connect to the ODBC data source in Power BI Desktop, there is a blog for your reference.

Thirdly, in Power BI Desktop, you are able to develop your own custom connector.

Regards,
Lydia

[IoT-Guy]

Hello Lydia,

thanks for your answer.

We are storing our data in a hadoop cluster (structured data) and as well in S3 (unstructured data). It's good to know, that I can connect multiple data sources. Your answer helped a lot.

Best,

Michael