Within Integration cloud we are moving to a model that allows us to trial new features without making them available to everyone. Everone runs the same codebase but feature flags control what is available to a specific instance. Why would we do this? For multiple reasons:
Each new feature is given a flag that is used to control its availability. For instance the flag for the small footprint OIC agent was oic.adapters.connectivity-agent.light-weight-agent. If this flag was enabled for a given OIC instance then they could download the lightweight connectivity agent. Other OIC instances running the same code but with the flag turned off would not offer the new agent.
Flags are controlled from a central system and can be updated in real time by Oracle development and operations. This means that feature flags can be turned on very quickly, and also if a problem occurs they can be disabled.
Feature flags have a lifecycle as illustrated below.
The different stages are:
You may see a product manager demo features on an instance that are not currently available, if using a production pod these may only be available to internal users. This is where we try things out internally before turning them on for any customers. Once we are happy with the feature internally we are ready to share it with selected customers and move the feature to Feature Controlled. Note that this change in stage does not require any code changes, it just alters our internal approval process to enable the feature. Read the complete article here.
For regular information on Oracle PaaS become a member in the SOA & BPM Partner Community for registration please visit www.oracle.com/goto/emea/soa (OPN account required) If you need support with your account please contact the Oracle Partner Business Center.