Integrationsruntimes

Integration Runtime (IR) is the compute infrastructure used by data integration services to perform activities such as data movement, data transformation, and pipeline execution. In platforms like Azure Data Factory and Azure Synapse Analytics, the IR acts as the execution engine that connects data stores, runs transformations, and orchestrates activities across environments.

There are three primary types of integration runtimes. Azure Integration Runtime is a cloud-based, managed environment that executes copy activities and data flows, scaling to meet workload demands and connecting to a wide range of cloud data stores and services. Self-hosted Integration Runtime is installed on on-premises machines or in a private network and provides connectivity to on-premises data sources, hybrid networks, or data sources that are not directly reachable from the cloud. Azure-SSIS Integration Runtime is a dedicated runtime in Azure for running SQL Server Integration Services (SSIS) packages in the cloud, enabling lift-and-shift of SSIS workloads to Azure.

Azure IR can be used for data movement and transformation in the cloud, including mapping data flows and cloud-to-cloud data movement. Self-hosted IR is chosen when data resides behind firewalls or in private networks, requiring private connectivity or on-premises access. Azure-SSIS IR is selected when existing SSIS packages need to be executed in the cloud, often to preserve existing ETL assets and logic.

Management of IR instances is performed within the data integration service, allowing multiple IRs per workspace, with considerations for security, network configuration, and licensing. Proper selection of IR type depends on data location, network topology, and the need to run SSIS packages or complex data flows.