New Features and Improvements
AtScale contains the following new features and improvements.
C2026.5 is a short-term support (STS) release. For more information, refer to the AtScale Product Lifecycle Support Policy in the Help Center.
C2026.5.0
Runtime permissions for model resources
You can now configure runtime permissions for individual resources (such as metrics and calculations) in your deployed models. Resource permissions give you finer-grained control over access to your model data than model permissions, which apply to the entire model.
Resource permissions are only available for metrics and calculations.
For information on configuring resource permissions, see Configuring Resource Permissions for Deployed Models.
ATSCALE-21023
Global environment variables
You can define environment variables at the global level using the new sml.environment.variables global setting. Variables specified in this setting apply to your entire AtScale environment, unlike those defined via .env files, which apply to a specific repository.
Additionally, AtScale now provides the /settings/sml-environment-variables API endpoint for retrieving all currently configured global environment variables.
For more information on configuring global environment variables, see Configuring Global Environment Variables. For information on the new API endpoint, refer to the AtScale Public API documentation.
ATSCALE-44047
Enhancements to partition-based incremental aggregate rebuilds
AtScale's partition-based incremental aggregate rebuild functionality has been enhanced to streamline partition configuration: Instead of setting a partition strategy to determine how partition keys are selected, you now define the partition key directly on the dataset, via the Dataset properties panel.
Partition-based incremental aggregate rebuilds are a Public Preview feature.
For more information, see About Incremental Rebuilds and Configuring Incremental Aggregate Rebuilds.
ATSCALE-41898
Server-side DAX language support
AtScale now supports server-side DAX, providing an alternative to MDX for writing calculation expressions. This functionality helps streamline the process of converting Microsoft Power BI workbooks to AtScale models.
Server-side DAX language support is a Public Preview feature.
For information on working with server-side DAX in AtScale, see Supported Server-Side DAX Language Elements.
ATSCALE-43583
New MDX functions
AtScale now supports the NULLIFZERO and NULLIFZERO(scalar) MDX functions. For more information, see NULLIFZERO and NULLIFZERO(scalar).
ATSCALE-44658, ATSCALE-45087
Configure dataset immutability in Design Center
You can now configure dataset immutability in Design Center using the new Immutable option in the Dataset properties panel. Previously, this could only be configured via SML.
Marking a dataset as immutable tells AtScale the following:
- Aggregates built from the dataset do not require a dataset-ready signal from the ETL pipeline before being released after a rebuild.
- The dataset is a rarely changing dimensional dataset that is safe to include in incremental aggregate rebuilds.
For more information on the Immutable option, see Adding Datasets and Configuring Incremental Aggregate Rebuilds.
ATSCALE-45715
Enable/disable aggregate usage via API
AtScale provides a new /user-settings/aggregate-usage API endpoint for enabling and disabling the use of system- and user-generated aggregates for the current user. This enables you to query the underlying tables. For more information, refer to the AtScale Public API documentation.
ATSCALE-45507
query.factless.ignoreIncidentalFilter settings now false by default
The query.factless.ignoreIncidentalFilter global and model settings are now false by default.
If you have not explicitly set query.factless.ignoreIncidentalFilter in your global or model settings, dimension-only filter queries will now be cross-filtered across fact tables by default. The system will create aggregates to accelerate this workload, but the model must contain a simple base metric (Sum, Min, Max, or Count) for the fact dataset that connects the dimensions that the BI tool user wishes to cross-filter. Avg, Distinct Count, and Stdev metrics will not trigger the aggregate acceleration behavior.
If you do not want to have this cross-filtering enabled, AtScale recommends setting query.factless.ignoreIncidentalFilter to true in either your model or global settings file.
For more information, see Query Settings (global level), Query Settings (model level), and About Factless Queries.
ATSCALE-45204
Configure additional CA certificates in the AtScale Helm chart
You can use the new global.atscale.extraCaCerts and global.atscale.existingExtraCaCertsSecret parameters in the AtScale Helm chart to configure additional CA certificates for your AtScale instance. This feature removes the previous requirement to merge multiple CAs into a single base64-encoded caCerts blob. It is useful when a client proxy routes cluster traffic using a certificate signed by a different Certificate Authority than the primary AtScale TLS certificate.
For information on configuring additional CA certificates using the new parameters, see Configuring Additional CA Certificates.
DEVOPS-7154