Add Support of OCI Vault into Helidon

[paulparkinson]

This is the same topic as #2395 "Add Support of Hashicorp Vault into Helidon" but specifically the OCI Vault service and, like that ER, we are researching best approaches to this particularly as there are a number of solutions and approaches in this area.

[tomas-langer]

It would be good to understand what features are desired for OCI vaults (ideally with a reference to an SDK function, or REST API).
After playing around with the SDK, it seems there is no way to retrieve the actual secret value from a Vault, which seems to be the real use case I can see here.

I do not think that creating/updating secrets is the usual use case (please let me know if this assumption is wrong).

[tomas-langer]

Found the getSecretBundle() method that can actually return secret content, looking into it.

[tomas-langer]

Notes for follow up

DONE

• Add CDI support for reactive interfaces, so this can be tested in an MP application
• Add blocking APIs + CDI support for injecting them
• Validate all modules exist in bom/pom.xml
• Add CDI examples
• Validate reactive examples work with native-image
• Update FeatureCatalog with the new modules
• Validate module-info.java exists for all modules (except for examples)
• Validate an example using module path with full module info for both SE and MP

TODO:

• OCI - instance principal security #2948 Implement Instance principal security (for service running on OCI nodes - works in a simple VM)
• OCI - Resource Principal security #2949 Implement Resource principal security (for functions)
• OCI Integration - documentation #2950 Add documentation
• OCI Integration - AOT #2951 Validate native image
• OCI Integration - AOT #2951 Update AOT documentation (SE and MP for the new modules)
• Tracing of OCI calls #2952 Check and document integration with Tracing
• Health Checks for OCI #2954 Check and document integration with HealthChecks (maybe add a health check for OCI?)
• Metrics for OCI calls #2953 Check and document integration with Metrics

[trentjeff]

Discussed with @tomas-langer , we can close this issue.