Forum Discussion
PowerBI Gen1 dataflow change Gateway
You're planning to migrate some Gen1 dataflows to a new Power BI Gateway cluster to reduce the load on an existing cluster that's been overwhelmed by traffic. A key concern is whether dataflows with incremental refresh will be affected—specifically, whether changing the gateway will force a full refresh instead of preserving the incremental behavior. Generally, changing the gateway itself does not automatically trigger a full refresh, as long as the new gateway is correctly configured to match the original connection (same server name, credentials, privacy level, and data source path). However, if there's any mismatch or the system perceives the data source as different—for example, if the connection string changes slightly—Power BI may consider it a new source and invalidate existing refresh partitions, resulting in a full refresh. To avoid this, ensure the new gateway cluster uses identical settings and test with a non-critical dataflow first. This will help verify that incremental refresh continues to function as expected without data loss or unexpected processing overhead.
You're planning to migrate some Gen1 dataflows to a new Power BI Gateway cluster to reduce the load on an existing cluster that's been overwhelmed by traffic. A key concern is whether dataflows with incremental refresh will be affected—specifically, whether changing the gateway will force a full refresh instead of preserving the incremental behavior. Generally, changing the gateway itself does not automatically trigger a full refresh, as long as the new gateway is correctly configured to match the original connection (same server name, credentials, privacy level, and data source path). However, if there's any mismatch or the system perceives the data source as different—for example, if the connection string changes slightly—Power BI may consider it a new source and invalidate existing refresh partitions, resulting in a full refresh. To avoid this, ensure the new gateway cluster uses identical settings and test with a non-critical dataflow first. This will help verify that incremental refresh continues to function as expected without data loss or unexpected processing overhead.
Thanks Poojara_D12 for clarifying it