Skip to main content

Release Notes for Self-Hosted Styra DAS

Self-Hosted Styra DAS 0.17.1 was released on July 17th, 2025.

Self-Hosted Environment Changes​

Individual control over Elasticsearch shards and indices for status, decisions, and activity​

New feature flags are available in this Self-Hosted release to allow more fine-grained control over the number of shards and index duration for agent status logs, decision logs, and activity logs when stored in Elasticsearch. For most Self-Hosted customers, the existing default configuration is sufficient. For customers with either very high daily decision volumes or very large decision sizes, using an index duration of less than 24 hours (the current default) could provide better Elasticsearch cluster performance and stability. Contact your account representative or create a support ticket to discuss whether customizing this configuration option can solve any existing Elasticsearch issues or decision log indexing lag.

Increased Elasticsearch HealthCheck timeouts​

The HealthCheck timeouts DAS uses for Elasticsearch have been increased to reduce scenarios where an Elasticsearch node is prematurely removed due to a longer startup time or longer response time, which can result in Elasticsearch cluster instability and reduced performance.

New Features and Changes​

Upgraded to OPA v1.5.1​

The internal version of OPA used by Styra DAS has been upgraded to OPA 1.5.1.

New UI Additions​

Added functionality for System-level decision export to the New UI when tenants are using Agentloader v2 to match the legacy UI functionality.

New UI free-form System discovery configuration editor​

For advanced users who may need additional customization of the discovery configuration served by DAS to OPA and Enterprise OPA, users can now use a free-form editor in the New UI to customize a System's discovery configuration. The discovery configuration editor can be found in the New UI in System settings under Agents > Advanced.

Git status in New UI System, Stack, and Library dashboard​

In the New UI dashboard for Systems, Stacks, and Libraries, the Git status is now shown in the dashboard status card and contributes to the System/Stack/Library status indicator on the dashboard. In cases of a Git sync error, the error will be displayed in the dashboard status card.

Improved New UI load times with many libraries​

In the New UI, for tenants with many libraries, UI load times of the System dashboard and the policy editor have been significantly improved.

Decision log indexing performance improvements​

Additional improvements have been made to the indexing of decision logs for decision search, which will result in indexing performance improvements for customers with high daily decision log volumes.

Improved OPA status processing performance​

For tenants with many OPA, Enterprise OPA, SLP, and Data Source agents, DAS is now able to process more agent status updates to provide close to real-time processing of those updates with the addition of multi-threading to the agentstatusstore service. If necessary, the AGENT_STATUS_THREADS feature flag can be used to adjust the number of threads from the default of 8.

New Prometheus metrics added​

The following new Prometheus metrics have been added to aid in monitoring DAS performance:

  • The styra_das_bundle_build_latency_seconds metric allows customers to monitor latency in bundle builds.
  • The styra_das_agent_status_compliance_system_latency_seconds and styra_das_agent_status_compliance_stack_latency_seconds metrics allow customers to monitor the latency in System and Stack compliance jobs.

SLP v0.10.5, CLI v0.13.4, and Relay Client v0.1.16​

SLP v0.10.5, Styra CLI v0.13.4, and Relay Client v0.1.16 have been released with dependency updates to resolve recently reported CVEs related to those dependencies.

Fixed Issues​

Agentloader v2 error for Kafka decision export​

In some cases during decision log export to Kafka, the channel to Kafka closed and Agentloader v2 would still try to send decision logs to export to the closed channel.

Omitted logs in activity log export when using Agentloader v2 with very high volumes​

For tenants using Agentloader v2 and with activity log export configured, in some cases logs could get omitted from the export in times of very high volumes of activity logs.

Omitted logs in activity log indexing when using Agentloader v2 with Elasticsearch indexing of activity logs​

For tenants using Agentloader v2 along with Elasticsearch indexing of activity logs enabled and the Activity v2 API enabled, some logs could get omitted from indexing into Elasticsearch.

Metrics export failure to DataDog or SignalFX Monitoring Integrations​

After configuring metrics export to DataDog or SignalFX in the Workspace-level Monitoring Integrations, the metrics-exporter service was unable to collect metrics due to a failure in communication with the Timeseries and Systems services, resulting in no metrics sent to the configured integration.

Incorrect metrics reported for decision indexing latency​

In some cases incorrect values were reported for the Prometheus metrics related to DAS decision indexing latency.

S3 decision export target error when using IAM role​

The incorrect trust policy was shown during the set up process of decision log export from DAS to S3 when using AWS IAM Role authentication, which resulted in failures to export decision logs to S3. The /v1/workspace/aws/config endpoint has also been updated for consistency to return the das_trust_role in the result.

Timeseries metrics graphs in loading status after API error​

If an API timeout or other error occurred when loading timeseries metrics graphs, the graphs would continue to show a loading state instead of showing an error message.

Workspace Usage tab error in New UI​

In the New UI Workspace Usage tab, there was an error when viewing the Usage tab if the decisions chart had no decision metrics data to display.

Monitoring graph tooltip shows N/A instead of 0​

In the monitoring graphs (e.g., decisions, violations), when the value in the graph for a metric was 0 the tooltip on hover incorrectly showed N/A.

New UI sidebar long loading times​

In the New UI, when viewing a System, Stack, or Library, the sidebar menu could at times take several seconds to load.

Decision metrics graph drop-off to 0 when waiting for new data​

When DAS is waiting for new decision metrics, decision graphs on the Workspace, System, and Stack dashboards may show a drop-off to 0 while waiting for current decision metrics data.

Some user roles in New UI unable to change password​

In the New UI, for tenants not using SSO, some user roles did not allow a user to change their own password.

Some access control buttons shown for WorkspaceViewer in New UI​

In the New UI User Access settings, users with the WorkspaceViewer were incorrectly shown active buttons to add a user, token, and SSO provider, however these users were correctly not able to create those resources.

WorkspaceViewer unable to view Slack and Terraform integration settings in New UI​

In the New UI, users with the WorkspaceViewer role were unable to view the configured Slack and Terraform integration details in the Workspace settings.

New UI Policy Builder fixes​

  • Policy Builder “New policy type from inputs” did not copy the inputs into the new policy.
  • Boolean type missing value in the Policy Builder create rule dialog.
  • Policy Builder UI crash when loading System with a policy type using a reserved Rego keyword.
  • Policy Builder graph showed undefined for rule conditions using contains.
  • Policy Builder error when a policy rule used the Rego count() built-in.