Skip to main content

Dear Experts,

 

Kindly guide me with your valuable comments on “How to connect from Power BI via API into IFS Cloud 23R2 to extract data for Reporting”? Any reference which explains the steps to set up this connectivity, please?

Please see https://docs.ifs.com/techdocs/23r2/050_reporting/450_self_hosted_bi_int/ .

In addition, you need to have the VPN in place from where you’re accessing in order to reach the database. Maybe the VPN was already built for migrations, but if not you can request it with a ‘Report something else’ ticket asking for a BI VPN connectivity (they’ll send a form to be filled in by your IT).

As you mention API, perhaps you are looking for a direct connection. Direct connections are not recommended, because BI activities produce a significant load on the IFS Oracle database causing performance issues for the regular usage of IFS Cloud.
The way IFS BI works is by transferring the data to an off-site BI database which can be queried by PowerBI. This solution is provided as supported by IFS, however, if required, it is technically possible to develop a non-supported solution using the same Access Views in IFS Cloud (details on Access Views through the menu of the link above).


@TatJayadP below is a blogpost written by @dsj and it’s a step by step tutorial on how to use a proxy in Azure to enable direct connection to IFS Cloud. I’ve built quite a few Power BI dashboards for my client using this easy method.

Creating an API Proxy for IFS Cloud using Azure APIM – DSJ's Blog (dsj23.me)


@Marcel.Ausan how is this performance-wise?
While small queries can surely be run, wouldn’t this expose the system to daytime large queries that should rather be run off-peak or off-site?
Depends on the usage of course, I’ve heard some customers have been successful with direct access for small queries even though it is not recommended. It’s a fine line and should be used with insight.


@SamiL agree. In my case I have created a few quick reports and via API Proxy I call the Quick Report projection from Power BI. It works extremely well, but also the dataset has less than 10K rows.

For sure, when complex ETLs are needed + historization, then this solution won’t cut it and then self-hosted BI would be the way to go. For less complex BI requirements I found direct connection via API proxy to work quite well.


Reply