Start a new topic

Global Config Question While running in Azure Kubernetes

Hello All, I just wanted to confirm that our global configuration is correct in our Semarchy xDM instances running in our Azure tenant/AKS Clusters.


active deployment:  dev-semarchy-admin.xyz.com

passive deployment: dev-semarchy.xyz.com


In the Semarchy xDM Global Configuration, should the "Server Base URL"(V_SERVER_BASE_URL) always be set to the "Active" service base url?

One quick follow-up question:  If you change the value of the "Server Base URL" in the Global Configuration user interface, do you need to restart your Active/Passive deployments in order for the change to take effect?


Hello Joe,


Which URL should the Server Base URL be set to?

The Server Base URL is a single setting stored in the repository. So it is the same for the active and passive instance (there is no separate value for the passive instance).

The setting is used to build links to the server in workflow notifications, and it is exposed to models as V_SERVER_BASE_URL.

It should therefore be the URL that the people clicking those links can actually reach.

This is not necessarily the "active" one by definition:

  • If end users work on the active deployment, set it to the active URL (dev-semarchy-admin.xyz.com).
  • If the active deployment is used for administration or design only, and users work on the other one, set it to the URL users open.
  • If the two deployments form a failover pair, where one serves and the other stands by, the setting needs to follow whichever one is serving. You can either update it as a step in your switchover procedure, or set it to a stable hostname that always routes to the live deployment, which avoids the manual step.


Is a restart needed after changing it?

No, it is not needed.

But please note that for variable values, they are cached per user session.

Everything is detailed in the documentation.

Users who are already logged in may see the previous value for up to 30 minutes after the change.

New sessions pick up the new value straight away, and refreshing the application applies it sooner.


Suggested check: after any change, trigger a workflow notification and confirm that the link shows the expected URL.


Best regards,

Alexia

Thank you, Alexia!  We stumbled slightly in our DEV environment during our xDM upgrades from 2024.1.1 to 2025.1.11(root deployment context change).  We had a leftover invalid base url setting in DEV that has since been corrected.  Good to know regarding the flexibility of the base url setting in our AKS config in case we choose to direct all traffic to the non-admin(passive) xDM replicas.


Regards,


Joe Patton

Login to post a comment