Isolate connectors so a failure in one does not cause full user ingestion failure

We believe it would be beneficial to isolate individual connectors so that a failure in one does not cause a full breakdown of user ingestion. For example, recently our Okta connector (used only for proactive account unlock) experienced a failure, and this halted all user ingestion, which is primarily done through our ServiceNow connector. Ideally, an Okta failure would not prevent all user ingestion from happening. This prevented all new hires from being ingested and also caused user attributes that had been updated in ServiceNow not to be updated in Moveworks.