How to Update Data Refresh Interval or Retain Spark Pool Association in Power Apps Azure Synapse Link

Question

We are extracting Dynamics CRM data using Power Apps Azure Synapse Link (ASL) into our Azure Storage Account. The ASL is configured with the required Storage Account, file system, and is linked to an Azure Synapse Analytics workspace. Using an Azure Synapse Spark pool, the data is successfully dumped as Parquet files, and the current data refresh interval is set to 30 minutes.

The issue we face is that once an ASL is created, the refresh interval cannot be modified directly. To attempt a change, we:

1. Exported the ASL as a solution into a zip folder.

Updated the IncrementalUpdateTimeInterval property in the customizations file within the solution.

Imported the solution back and then re-imported the ASL.

After importing:

The ASL did not pick up the Spark pool details.

The data refresh interval was not visible.

Data is now being dumped as CSV instead of Parquet, since the Spark pool association is lost.

Questions:

Is there a supported or workaround method to update the data refresh interval for an existing ASL without unlinking and recreating it?

How can we ensure that Spark pool properties are retained when importing an ASL solution so that it continues to write Parquet files?

Any guidance or best practices for managing these properties in Power Apps ASL would be greatly appreciated.

Answer 1

Hi @Srimanikanta Pallela
Thanks for explaining the scenario clearly.

Currently, in Power Apps Azure Synapse Link (ASL), the data refresh interval cannot be changed after the link is created. Updating the solution files manually (like modifying IncrementalUpdateTimeInterval) is not a supported approach, so the system will not apply those changes.

Because of this, the recommended and supported method today is to recreate the Synapse Link with the required refresh interval.

Regarding the Spark pool issue - Spark pool configuration is not retained when exporting/importing the ASL as a solution. After re-import, the link behaves like a storage-only link, which is why the data is now landing as CSV instead of Parquet. To continue writing Parquet files, the Synapse workspace and Spark pool must be configured again during link creation.

Best practice:

• Decide refresh interval before creating the link
• Recreate the link if interval needs to change
• Reconfigure Spark pool after recreation
• Avoid modifying solution files manually for ASL settings

If needed, you can share your current setup and we can suggest the safest recreation approach with minimal impact.

Hope this helps. Let us know if you have further questions.

Answer 2

Azure Synapse Link for Dataverse treats the Spark pool configuration and the Delta/Parquet conversion settings as part of the link profile configuration, not as solution-customizable metadata. The behavior described (Spark pool association lost and data reverting to CSV) is consistent with how profiles are created and managed.

From the available information:

1. Azure Synapse Link profiles that use Delta Lake/Parquet rely on an explicitly configured Spark pool in the associated Synapse workspace. The Spark pool is used for Delta conversion and is bound to the profile at creation time.
2. Delta conversion frequency (data refresh interval) is configured as part of the profile and directly affects Spark pool usage and cost. Lower intervals consume more Spark compute and can increase cost.
3. When a profile is created without a Spark workspace/pool, or when the Spark pool is not configured, data is written as CSV rather than Delta/Parquet.
4. The documented way to change Spark-related characteristics for a profile is effectively to create or assign an appropriate Spark pool and, when upgrading runtimes, to create a new Spark pool and then upgrade the Synapse Link profile to use it (in-place upgrade). This is done through the product’s upgrade flow, not by editing solution XML.

Based on this, the following conclusions apply to the questions:

1. Updating the data refresh interval for an existing Azure Synapse Link profile by editing and re‑importing a solution is not a supported method. The interval is a runtime configuration tied to the profile and Spark usage; the documentation only describes setting it at profile creation time and adjusting Spark pool size to meet refresh needs. No supported mechanism is documented to change the interval post‑creation via solution customization.
2. Spark pool association is not preserved or controlled via solution export/import. Spark pool configuration is done in the Azure Synapse Link profile UI and depends on an existing Spark pool in the Synapse workspace. When the profile is configured without a Spark pool, or when the Spark pool is unavailable, the service falls back to non‑Delta export (CSV). There is no documented way to “carry” Spark pool properties through solution import so that they are automatically re‑bound.
3. Recommended practice for managing refresh and Spark behavior is:
   • Create and manage Spark pools directly in the Synapse workspace (for example, via + New Apache Spark pool in the workspace) and size them appropriately for the chosen refresh interval.
   • Use the Azure Synapse Link configuration/upgrade flows (including the in‑place upgrade process when changing Spark runtime versions) to bind profiles to the correct Spark pool and adjust pool size, rather than attempting to modify profile internals via solution XML.
   • If a different refresh interval is required and the product UI does not allow changing it on an existing profile, the supported path is to create a new Azure Synapse Link profile with the desired interval and Spark configuration, then cut over downstream processes to the new export.

No supported workaround is documented that would allow changing the refresh interval or retaining Spark pool association solely via solution export/import or direct manipulation of the IncrementalUpdateTimeInterval property.

---

References:

• Azure Synapse Link for Dataverse troubleshooting guide
• Access audit data using Azure Synapse Link for Dataverse and Power BI
• Choose finance and operations data in Azure Synapse Link for Dataverse
• Transform Azure Synapse Link for Dataverse data with Apache Spark
• Export Dataverse data in Delta Lake format
• What is Azure Synapse Link for Dataverse?
• Azure Synapse runtimes