You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A *NO-OP *change goes through the pipeline and into an environment without downtime
Who’s it for
As a dotCMS developer, I can be confident that my local, sandbox, qa, and production environments are consistent, so that I can separate concerns and interact with upgrade and running behaviour.
As a dotCMS customer, I want to apply a config change without downtime
What does it do
Allow a developer to interact with upgrades/deployments
Test and interact with the init container
Test that a no-op change is indeed zero downtime
reproducibility between local and sandbox environments (via Helm charts)
Infra config
Envvars
…etc
A config change should allow for restart without downtime.
What does it NOT do
DB and API changes would work (via init container) but with downtime
It does not guarantee that something is broken beyond dotCMS server starting up
Allow for anything other than starter starter as dotCMS
Solve the need to manually update the image name/version in the values file to rollout a new version.
What is required for this to be available
Tomcat redis sessions
Current -> current
What sprint size things do we need to do to get there and that we can demo?
Tomcat redis sessions for sandboxes
dotCMS image becomes the init container
Connect helm-chart into the sandbox (sandboxes get additional functionality via helm-charts and values.yaml file)
Replace scripts in sandbox tooling with helm-charts
Versioning for helm charts
Clean up the logs with probe where it throws exception each time it’s hit
Helm charts for dotCMS cloud customers on current
Important
The end result here is that the 8 customers on current are running using the same helm charts we're using locally and in sandbox
The text was updated successfully, but these errors were encountered:
Infrastructure and Config promotion
Important
The end result here is that the 8 customers on current are running using the same helm charts we're using locally and in sandbox
The text was updated successfully, but these errors were encountered: