fbcideas_migusr
11 years agoNew Member
Status:
Planned
Parallel processing of multiple partitions in the same table in a Tabular model
Currently the SSAS Tabular engine doesn't process partitions in the same table in parallel when loading data into the model. This is a major limiting factor on very large models.
1 Comment
- fbcideas_migusrNew MemberStatus added:Planned
Recent ideas
Prevent Overwriting Subscription Parameters When Adding Report Parameters
Summary: Currently, when a paginated report is republished with a new parameter, making any subscription changes (e.g., adding or editing one) overwrites all existing subscriptions' parameter values...andrew41 hour agoNew MemberNew825Views8likes2CommentsAllow Mirror Database to use Variable Library, not just Deployment Rules
Currently, the source connection for Mirrored DBs is managed through Deployment Rules in data pipelines. There is no way to use Variable Libraries with Mirrored DB, like you can do in CopyJob or Dat...hedeencharles2 hours agoAdvocate INew428Views11likes1CommentAzure Maps Power BI - Support Dynamic Geometry Data from Dataset Columns
Problem The Azure Maps Reference Layer currently relies on file uploads or URL-based geometry sources (GeoJSON, KML, etc.). In our enterprise use case, geometry data is private, changes daily, and m...jurner4 hours agoNew MemberNew24Views1like0Comments