The Delta table uses the 'v2Checkpoint' feature, which is not supported : Azure Databricks to Fabric

[MarouaneB]

Hello everyone,

I’m trying to shortcut/mirror several Delta tables from Azure Databricks to Microsoft Fabric. These tables have the checkpoint feature enabled, and Fabric fails with the following error:

Error: The Delta table uses the v2Checkpoint feature, which is not supported.

Error code: UnsupportedV2Checkpoint

Exception type: Microsoft.DeltaLogParser.Exceptions.DeltaLogParserUserException

Questions:

How can I disable the v2Checkpoint feature on existing Databricks Delta tables so I can bring them into Fabric without compromising data ?

Why are some tables created with these features enabled while others are not? (Same workspace, similar pipelines.)

Details (from Databricks catalog for the tables that fail):

delta:
checkpointPolicy: "v2"
columnMapping.maxColumnId: "16"
columnMapping.mode: "name"
enableDeletionVectors: "true"
enableRowTracking: "true"
feature.clustering: "supported"
feature.columnMapping: "supported"
feature.deletionVectors: "supported"
feature.domainMetadata: "supported"
feature.identityColumns: "supported"
feature.invariants: "supported"
feature.rowTracking: "supported"
feature.v2Checkpoint: "supported"
lastCommitTimestamp: "1760449400000"
lastUpdateVersion: "302"
minReaderVersion: "3"
minWriterVersion: "7"
rowTracking.materializedRowCommitVersionColumnName: "_row-commit-version-col-dbf6ae64-37a5-49ee-9a03-d36feb509e0a"
rowTracking.materializedRowIdColumnName: "_row-id-col-0a7bda7f-5e15-4595-a6aa-c92029bac0e6"

Any guidance (workarounds, migration steps, or best practices to avoid enabling unsupported features) would be greatly appreciated. Thanks!

[MarouaneB]

Hello, This is clear, thank you for your help. 

[v-pgoloju]

Hi @MarouaneB,

Thank you for reaching out to the Microsoft Fabric Forum Community, and special thanks to @Thomaslleblanc for prompt and helpful response.

Just following up to see if the Response provided by community members were helpful in addressing the issue. if the issue still persists Feel free to reach out if you need any further clarification or assistance.

Best regards,
Prasanna Kumar

[Thomaslleblanc · Accepted solution]

• If the table is already created with v2Checkpoint, consider rewriting or exporting the data into a new Delta table using the classic checkpoint format (JSON).

You can enforce classic checkpoints in Databricks by setting:
spark.conf.set("spark.databricks.delta.checkpoint.writeFormat", "json")