cancel
Showing results for 
Search instead for 
Did you mean: 
Community Platform Discussions
Connect with fellow community members to discuss general topics related to the Databricks platform, industry trends, and best practices. Share experiences, ask questions, and foster collaboration within the community.
cancel
Showing results for 
Search instead for 
Did you mean: 

Databricks Materialized View - DLT Serverless Incremental

ismaelhenzel
Contributor

I'm currently working with Delta Live Tables, utilizing materialized views and serverless computing. While testing the incremental updates of materialized views, I've observed that deleting a record from the source table triggers a complete refresh of the materialized view. Upon reviewing the materialized view logs, I discovered this behavior:


message: Flow 'teste_novo_2' has been planned in DLT to be executed as COMPLETE_RECOMPUTE. Another option is available:ROW_BASED. COMPLETE_RECOMPUTE was chosen in the current run for its optimal performance.

details: 

{"planning_information":{"technique_information":[{"incrementalization_issues":[{"issue_type":"INCREMENTAL_PLAN_REJECTED_BY_COST_MODEL","prevent_incrementalization":true,"cost_model_rejection_subtype":"CHANGESET_SIZE_THRESHOLD_EXCEEDED"}]},{"maintenance_type":"MAINTENANCE_TYPE_COMPLETE_RECOMPUTE","is_chosen":true,"is_applicable":true,"cost":4.75717377E8},{"maintenance_type":"MAINTENANCE_TYPE_ROW_BASED","is_chosen":false,"is_applicable":true,"cost":1.18965781E8}],"source_table_information":[{"table_name":"`mul_dev_tests`.`dlt_managed`.`tabela_teste_4`","table_id":"c7147455-60dd-442b-924e-ca103c82f026","full_size":4.79009539E8,"change_size":1.19579857E8,"is_size_after_pruning":true,"is_row_id_enabled":true,"is_cdf_enabled":true,"is_deletion_vector_enabled":true,"is_change_from_legacy_cdf":false}],"target_table_information":{"table_name":"`mul_dev_tests`.`dlt_managed`.`teste_novo_2`","table_id":"bfc03f09-286d-4da1-bf71-32f80ebf1739","full_size":3.42395821E8,"is_row_id_enabled":true,"is_cdf_enabled":true,"is_deletion_vector_enabled":true}}}

What does the CHANGESET_SIZE_THRESHOLD_EXCEEDED error mean? Is there a resource that lists all incrementalization errors? Additionally, is there a method to handle deletions incrementally in materialized views?

obs: i tested delete only one line, and 3k lines, the same happens.

2 REPLIES 2

lucassvrielink
New Contributor

Same problem here

ismaelhenzel
Contributor

EDIT: My Delta Lake table contains 136 columns. I initially tested with fewer columns, and both the updates and deletes were applied incrementally without issues. Specifically, I tested with 34 columns, and everything worked fine. However, when I increased the number of columns to 40, I encountered the error mentioned above.

Connect with Databricks Users in Your Area

Join a Regional User Group to connect with local Databricks users. Events will be happening in your city, and you won’t want to miss the chance to attend and share knowledge.

If there isn’t a group near you, start one and help create a community that brings people together.

Request a New Group