Bug: Appsemble pod restarts under critical load

When running the complex and long duration refresh function of the SG Nummerplan app the Appsemble pod is prone to restart halfway through the requests. This results in empty tables and broken functionality. We need to find out why this happens and if it can be fixed with the current feature set.

The refresh function consists of a bulk delete of 1600 and 2400 resources a looped fetch of 1600 MS Graph API results (100 rows per request) and a matching of 2400 items in a data loader. So I can imagine why it is a heavy load for the platform. Either we improve resource handling with issue #1731 or we need to see why it crashes and if we can improve it.