Addressed Issues

canvas improved handling of file fields from python action previously, when playbook schemas were retrieved via an http action and passed into a python action, an error occurred if a file was defined as an object in the schema the issue has been resolved and python actions now work correctly with file fields in playbook schemas improved error messaging for python action outputs when a python action produces outputs that aren’t json serializable, the platform now provides a clearer error message this update replaces the previous generic “json serialization error” with a more descriptive message explaining that python action outputs must be json serializable fixed an issue where data was not accessible from parallel actions nested within other parallel actions applications and dashboards fixed an issue in application workflows where configuring a condition with a user/group field caused selected values to disappear after saving the workflow editor now correctly retains selected users and groups, ensuring values persist after save actions when defining workflow conditions fixed an issue where the application history endpoint could fail when returning large responses due to missing pagination the endpoint is now paginated to prevent memory overloads and improve performance when retrieving applications with extensive revision histories fixed an issue where special characters such as “&” were incorrectly displayed as their html entity equivalents (for example, \&) in dashboard reports with this fix, all special characters—such as ampersands and commas—now render correctly across all dashboard and report views fixed an issue where the date picker tooltip displayed the same value for both the start and end dates after a page refresh fixed an issue where the first navigation attempt to dashboards across tenants sometimes displayed a “system error occurred” message due to failed statistics calls fixed an issue where selecting a relative date range (such as “this week” or “this month”) multiple times in the date picker converted it into a fixed timestamp, causing the same start and end dates to appear the relative date range now remains consistent, ensuring that filters like “this month” continue to function dynamically as intended fixed an issue where the selected date range in the calendar was not retained after a page refresh the highlighted range now correctly remains visible as long as the start and end date values are still present fixed an issue where custom relative date ranges were not consistently supporting all documented time units and syntax, causing valid expressions (for example, now 5d to now+5d) to return invalid results the filter now correctly evaluates all supported time units — minutes, hours, days, weeks, months, quarters, and years — for both past and future ranges, ensuring accurate and reliable date based filtering reports and records fixed an issue where the advanced pie chart component displayed incorrect decimal percentage values, causing totals to appear lower or higher than 100% the chart now accurately calculates and displays percentage values so that the total always sums to 100% fixed an issue where long labels in the sankey chart overflowed outside the chart boundaries, affecting dashboard readability the chart now automatically truncates lengthy labels with an ellipsis (…) based on available width, preventing overflow and improving visual consistency fixed an issue where user presence icons in the record view did not update correctly when a user closed their browser tab, window, or navigated to another website previously, the icons for inactive users remained visible until the record was refreshed or all users exited the view the system now accurately updates user presence in real time, ensuring that icons appear and disappear based on active participation in the record fixed an issue where opening a record from global search or search results sometimes displayed outdated data from a previously opened record the record view now correctly refreshes with the selected record’s latest data without requiring a manual page reload user management users logging out via the api now have more control with a new optional parameter called globallogout global logout globallogout=true – logs the user out of all active sessions across all browsers and devices session logout globallogout= false – logs the user out only from the current session turbine ui continues to support current session logout , allowing users to remain logged in on other browsers or devices fixed an issue where users were redirected back to an unauthorized page after re logging into the application users are now correctly taken to their default landing page upon login fixed an issue where non admin users with granted read access were unable to open assigned applications and encountered unauthorized errors non admin users can now successfully access applications they have permission to view without encountering access errors tenant filter fixed an issue where filtering tenants on the roles page or in configuration manager returned unintended results the filter now correctly distinguishes between tenant names and tenant ids when a search term matches both a tenant name and a tenant id (for example, if both contain “12”), the filter returns both tenants to ensure complete and consistent results resolved an issue where the tenant search input field was not fully visible in the firefox browser the text in the tenant dropdown’s search field now displays correctly, ensuring users can view and enter search terms without any clipping or cutoff ssp import/export addressed an issue where the same workflow could be assigned to correlation fields in multiple applications this could result in orphaned orchestration tasks if one application was deleted going forward, the system prevents assigning a flow to more than one application’s correlation fields, and cleans up any tasks linked to an application when it is removed this means your orchestration tasks and application records stay aligned and accurate resolved an issue where exporting a playbook would fail after deleting an associated application previously, leftover orchestration task references caused the export process to error out with this fix, all orchestration tasks tied exclusively to a deleted application are removed, allowing smooth playbook exports going forward fixed an issue where connectors and http actions could retain links to deleted assets, causing import/export and workspace errors now, when an asset is deleted or renamed, its references in playbooks are cleaned up or updated—helping keep your system consistent and error free resolved an issue where references to an application in record actions within playbooks weren’t being cleaned up when an application was renamed or deleted renaming an application now updates its references automatically, and deleting an application removes its references entirely keeping your playbooks clean and export ready fixed an issue where playbooks containing script or transformation actions with leading newlines caused ssp import failures newlines at the beginning of these actions are now automatically removed when saving the configuration, ensuring smooth playbook export and import operations resolved an issue where using the v2/playbook api endpoint could create orphaned playbooks when saving after the original playbook was deleted the endpoint now properly validates the solutionid field and prevents creation of invalid or unlinked playbooks, ensuring consistent playbook management and preventing export failures fixed an issue where playbooks appeared as disabled in the import summary even after being successfully imported in an enabled state the ui now correctly updates to reflect the playbook’s enabled status, removing any misleading “disabled” messages after import fixed a ui issue where the enable playbooks button continued to appear after importing an ssp, even when the “enable playbooks” toggle was switched on before import the button now correctly remains hidden after a successful import, ensuring a consistent and accurate user experience remote agents the remote agent installer now includes preflight checks for operating system limits (ulimits) these checks alert users if system settings are too low to run multiple concurrent actions reliably if a limit is detected that could impact remote agent performance, the installer will surface a warning so user can adjust settings before proceeding the remote agent update process is now more robust during an update, the script will start the new agent containers and verify they’re running correctly before removing the old ones if the new agents fail to start or don’t pass the health check, the old agents will automatically restart, ensuring continuity and easy recovery remote agents no longer support cgroups v1 this change aligns with upcoming podman updates, which are deprecating cgroups v1 to ensure compatibility and reliable deployments, all installations must use cgroups v2 if a system is configured with cgroups v1, the installation script will now show a clear error and prevent installation