Alm_InvalidRequest_PurgeRequired Failure

[v-keshavjain]

Library Version

0.1.19

What is the problem?

When deleted some tables or columns in a Semantic Model, getting the following error while deployment:

Error Message:
Operation failed. Error Code: Alm_InvalidRequest_PurgeRequired. Error Message: Dataset changes will cause data deletion.

Is there an ideal solution?

Not sure about the solution, but found a similar issue addressed in Fabric Community: https://community.fabric.microsoft.com/t5/Developer/Error-quot-LRO-API-Error-Alm-InvalidRequest-PurgeRequired/m-p/4651090

Additional context, screenshots, logs, error output, etc

No response

[shirasassoon]

v-keshavjain Thanks for raising this. If you can give me some more context, what approach did you take to remove the tables/columns, did you do it in Tabular Editor or in a different way?

I tried to repro with a basic semantic model, but did not encounter this issue. If you have more details on the model that is throwing this error that would help my investigation.

[shirasassoon]

v-keshavjain Bump - is this still a bug on your end?

[luanlopez]

I'm experiencing the same issue. I'm using ms-fabric-cli to deploy my reports and semantic models, and occasionally I encounter the following error:

x import: [LongRunningOperationFailed] Operation failed: {'errorCode': 'Alm_InvalidRequest_PurgeRequired', 'message': 'Dataset X changes will cause data deletion.'}

[shirasassoon]

Luan Lopes da Silva (@luanlopez) Please check out this link and let me know if this helps mitigate.

[JacobKnightley]

Confirmed with product that is a known issue and limitation of the API. Unfortunately, at this time. The only option is to physically make the change in the UI. Then you will be unblocked.

[zzKool]

Confirmed with product that is a known issue and limitation of the API. Unfortunately, at this time. The only option is to physically make the change in the UI. Then you will be unblocked.

Hi Jacob Knightley (@JacobKnightley),
We also get this error pretty often.
What do you mean with "to physically make the change in the UI" ?
Til now we delete the model and redeploy it. But this is requires to reprocess the model and remap the reports to the new datasetID.

[JeremyDel]

Wouldn't it be possible to add a purge step if semantic data model is different ?

[shirasassoon]

Wouldn't it be possible to add a purge step if semantic data model is different ?

We have an active issue to address a workaround for this #574.

[shirasassoon]

Closing this issue as it is being tracked elsewhere.