Manage on-premises data gateway high-availability clusters and load balancing
We recently revised the on-premises data gateway docs. We split them into content that's specific to Power BI and general content that applies to all services that the gateway supports. You're currently in the general content. To provide feedback on this article, or the overall gateway docs experience, scroll to the bottom of the article.
You can use an on-premises data gateway cluster to avoid single points of failure and to load balance traffic across gateways in a cluster.
High-availability clusters for an on-premises data gateway
You can create high-availability clusters of gateway installations. The clusters help ensure that your organization can access on-premises data resources from cloud services like Power BI and PowerApps. Gateway admins use such clusters to avoid single points of failure when accessing on-premises data resources.
The gateway cloud service always uses the primary gateway in a cluster unless that gateway is not available. In that case, the service switches to the next available gateway in the cluster.
Manage a gateway cluster
After you create a cluster of two or more gateways, all gateway management operations apply to every gateway in the cluster. These operations include granting administrative permissions to a gateway and adding data sources or connections.
For example, when admins select Manage gateways in Power BI, they see the list of registered clusters or individual gateways. But, they don't see the individual gateway instances that are members of the cluster.
All requests are routed to the primary instance of a gateway cluster. If the primary gateway instance isn't online, the request is routed to another gateway instance in the cluster.
Load balance across gateways in a cluster
You can choose to let traffic be distributed evenly across gateways in a cluster. By default, the selection of a gateway during load balancing is random.
For example, to provide load balancing from the Power BI service, select the gear icon in the upper-right corner, then select Manage gateways. Next, select Distribute requests across all active gateways in this cluster.