When OPAL starts it bring the agent to the baseline data state(as indicated by data-sources-config) and follows to update it with incoming events. Read more here: https://docs.opal.ac/FAQ#what-is-the-difference-between-initial-data-load-baseline-and-follow-on-updates-

Note that any disconnection from the OPAL server, the client assumes it's out of sync, and reloads the entire state from scratch (yes that means your data-sources-config should always point to a full current state)

There are new features in development (e.g. data tracking) that would give OPAL more visibility into what things changed (still without reading their content) to provide more flexibility here