In most situations where the Power BI desktop report refreshes fine but the Power BI Service report fails to refresh, the probable causes stem from how the two environments manage data connections, provisioned credentials, or setup configurations. Below is a systematic approach to resolving such refresh failures.
1. Check Data Source Credentials and Gateways
Most instances of Power BI Service refresh-related issues often come as a result of credentials. While in the Organisational Workspace in Power BI Service, click on the Settings icon and select the relevant dataset. Next, navigate to the ‘Data source credentials’ section and confirm the entered credentials. Ensure that the entered credentials are those expected and that the appropriate consent has been given to any more permissions needed. Also, if on-premise data sources are used, make sure you have the upgrade installed and properly configured and that the on-premise data gateway is actually up and running. The Power BI Service specifically makes use of an on-premises data gateway to access data sources. Thus, ensure the settings and configuration of the on-premises data gateway for the data source are the same and make changes if necessary.
2. Examine Query Dependencies and Cloud Availability
The Power BI Service depends on cloud technology. Therefore, some of the data sources or transformations that are performed on a client machine may not work the same way when accessed online. For example, the use of custom connectors, some M language transformations, or connecting to unsupported data locations will not be possible. Where in your report there are data sources confined to the on-premises data gateways, or where there are specific operations that require dependent local data, it is advisable to either consider changing these queries to cloud-friendly ones or find other cloud-based solutions.
3. Consider if there are any restrictions regarding data models or file sizes
Power BI Service imposes file size restrictions that may be hidden in Power BI Desktop. For instance, the service is limited to 1 GB per dataset unless Power BI Premium is used where larger datasets are accepted. Furthermore, sophisticated calculations, cases, or large-size tables can also cause excessive pea of the system resources when performing refresh activities. The reliability of the refresh process can be enhanced by simplifying the data model, decreasing the table sizes, or dividing the model into several smaller models. Once modifications have been made, repost your report and track the process of the refresh for the resolution if there are any concerns.