Do I need a gateway if all my data is already in the cloud?

No. If every data source you connect to, such as a cloud SQL database, Dataverse, or a SharePoint Online list, is already reachable over the internet with proper authentication, Power BI can connect and refresh directly without a gateway. Gateways are only needed for sources that live on your own network or behind a firewall, such as an on-premises SQL Server or a local file server.

What is the difference between a personal gateway and an on-premises data gateway?

The on-premises data gateway (standard mode) supports scheduled refresh, DirectQuery, and can be shared by an entire team, with an administrator managing data sources and permissions centrally. The personal gateway is limited to scheduled refresh only, cannot be shared with other users, and is meant for a single person connecting to their own local files or databases. Microsoft recommends the standard gateway for any organisational use.

How does the gateway keep my credentials secure?

The gateway encrypts data source credentials before they leave your network, using a recovery key you set during installation that never leaves your control. Data itself is encrypted in transit between the gateway and the Power BI Service using HTTPS, and the gateway only opens outbound connections, meaning you do not need to open inbound firewall ports for it to work.

Can multiple reports share the same gateway?

Yes. A single on-premises data gateway installation can serve as the connection point for many datasets and many data sources at once. Organisations typically install one or two gateways in cluster mode for redundancy, then register all their on-premises data sources against it, rather than installing a separate gateway for every report.

What Is a Power BI Gateway?

A Power BI gateway is a piece of software that lets the cloud-based Power BI Service securely reach data sitting on your own network, such as an on-premises database or a local file share.

If a dataset refresh in Power BI fails with a vague connection error and your data lives on a server inside your own office or on a private cloud VM, the missing piece is almost always a gateway.

Why the cloud cannot just reach your database directly

The Power BI Service runs entirely in Microsoft’s cloud. Your on-premises SQL Server, file share, or SAP system typically sits behind a corporate firewall with no direct internet exposure, and for good reason: you do not want a database listening for connections from the open internet. The gateway solves this by running inside your network and initiating an outbound connection to the Power BI Service. Because the connection is always outbound, you never have to punch an inbound hole in your firewall, which is one of the main reasons gateways are considered a safe way to bridge cloud and on-premises data.

Standard mode versus personal mode

Microsoft ships two gateway modes. The on-premises data gateway (standard mode) is designed for organisations: it can be shared across an entire team, supports scheduled refresh and DirectQuery, and lets an administrator manage which data sources are registered and who is allowed to use them. The personal gateway is scoped to a single user, supports scheduled refresh only, and cannot be shared. For any business scenario beyond a single person automating their own report, the standard gateway is the right choice.

What the gateway actually does during a refresh

When a scheduled refresh runs, the Power BI Service sends the query to the gateway, the gateway executes it against the on-premises source using the credentials you configured, and the results travel back through the encrypted channel to the cloud. For DirectQuery or Live Connection reports, this same path is used every time a user interacts with a visual, so gateway performance directly affects how responsive the report feels. This is also why gateway placement matters: install it on a machine with a fast, stable connection to the data source, ideally on the same network segment.

High availability with gateway clusters

A single gateway installation is a single point of failure: if that machine is rebooted or goes offline, every scheduled refresh depending on it will fail. Organisations that rely on gateways for business-critical dashboards typically set up a gateway cluster, which is two or more gateway installations registered under the same logical gateway name. Power BI automatically routes requests to whichever member of the cluster is available, so a single machine going down does not interrupt refreshes.

Registering data sources and managing access

Once installed, an administrator registers each data source (a specific server and database, for example) against the gateway in the Power BI admin portal, and grants specific users or groups permission to use that connection when building datasets. This is a useful separation of concerns: the person who knows the database credentials configures the connection once, and report authors across the organisation can then build against it without ever seeing the actual password.

Where the gateway fits in a larger Power BI setup

Gateways typically show up alongside other pieces of a mature Power BI environment. If you are publishing curated content to business users through a Power BI app, the underlying dataset refresh very often depends on a gateway if any source data is on-premises. And if you are managing separate development, test, and production environments, gateway configuration is one of the things you need to account for consistently across stages; see our article on what a Power BI deployment pipeline is for how that process works.

Our tip: Install the gateway on a dedicated server rather than someone’s laptop. A gateway tied to a personal machine that gets shut down overnight or taken home will quietly break every refresh scheduled outside office hours.