The documentation for ACE Configuration CRDs recommends no longer using spec.contents to store Base64 encoded data when the Configuration does not contain secrets like for instance odbc or policy types.
|
spec.contents
|
If the type of configuration that you are creating does not contain secrets, spec.contents is used to store the Base64-encoded value (of a configuration file) that you specify in spec.data. The value is automatically transferred from spec.data to spec.contents after you create the configuration.
Note: In 11.0.0.9-r3 or earlier versions, spec.contents is used for the manual input of Base64-encoded content. Manual input in this field is no longer recommended in later versions.
|
https://www.ibm.com/docs/en/app-connect/container?topic=resources-configuration-reference
Using the spec.data field when just applying the Configuration yaml manually or as part of a pipeline works fine. But it does not work in a gitops setup where ArgoCD is used to implement gitops. This is because the Configuration definition in git uses the spec.data field and the actual configuration in the OpenShift cluster uses spec.contents field. ArgoCD will therefore report the resource as being OutOfSync.
For the time being we will continue using the spec.contents field. But we would really like to know the rationale behind the recommendation that the spec.contents field should no longer be used!
Note that this is not an issue for Configurations containing secrets because there it is possible to reference a secret instead of adding the Base64 encoded data directly. I can imagine using a similar approach, using configmaps instead of secrets, for Configurations not containing secrets in a future release. But that is pure speculation.
------------------------------
Ron Peereboom
------------------------------