209 - Max Verstappen

Last modified by Carlijn Kokkeler on 2024/05/22 13:37

In the last sprint cycle, we focused on improving the speed of the deployment plan and making the next generation architecture the default. On top of that, we have worked to release new cloud templates for Docker Single Lane and Docker Double Lane, to introduce faster machine boot up and improved auto-healing. Moreover, we processed several feedback items and did some bug fixes.

Please find out more on our Release blog.

New features

  • Deploy - Releases: We added the ability to see all unused properties of an environment. This can be viewed by clicking the three dots in the Create phase section. Here it is also possible to quickly delete all or a selection of those unused properties.
  • Deploy - Cloud Templates: New Cloud Templates are available for docker single and double environments to introduce faster machine boot up and improved auto-healing. Please check out the cloud template release notes for more information.
  • All - The next generation architecture is the default now. New models will use this generation as default.

Minor changes

  • Deploy - Releases: We sped up the process of preparing runtime images in executing the deployment plan. The step in the deployment plan concerning the preparation of runtime images will now be executed more quickly and reliably.
  • Manage - Monitoring: The queue browser now displays milliseconds in the timestamps.
  • Manage - Alerting: Alerts will only be generated for queues that are created and managed by the eMagiz JMS server.

Bug fixes

  • Create - Flow Designer: An issue has been fixed, where, previously, the connection line for drawing channels did not work well after the user scrolls or zooms in/out on the canvas.
  • Deploy - Deployment plan: We fixed an issue of the release preparation step taking longer than necessary which is caused by the machine deployment steps to be executed when they could have been skipped.
  • Manage - Monitoring - We fixed a rare case where a runtime either (1) could not start, or (2) logging, errors, and metrics could not be seen in the portal for a model that had just migrated to the next generation architecture.