Understanding integration runtimes
IRs are responsible for all the compute jobs in Synapse pipelines or Azure Data Factory. When it comes to data movement as well as running the data flow activities (refer to the preceding Understanding mapping flows and Understanding wrangling flows sections), additionally, the IRs invoke activities and take care that they are run on their respective compute environments. Running SQL Server Integration Services (SSIS) packages is another responsibility that IRs handle.
There are different types of IRs available in Synapse pipelines/Azure Data Factory:
- Azure IR: This will take care of data movement activities such as data copy, control all activity runs, and run data flows (mapping and wrangling flows).
- Self-hosted IR: This must be installed in any on-premises network to "see" data sources that are loaded from on-premises to the cloud. They act as a data gateway and are responsible for the data movement from on-premises to...
