Getting “Datasource not found on datastore server” after “Finished scanning inbox” for some milliseconds.

[fm3]

Context: https://scm.slack.com/archives/C143WEU77/p1747902572736559?thread_ts=1747211202.420449&cid=C143WEU77

I originally thought the datastore must have restarted (possibly due to OOM) but it looks like this “initializing” state happens for a short period of time after each “sanning inbox”. We should have a look if we can get this to be safe / atomic.

Current code just does

      _ = removeAll()
      _ = dataSources.foreach(dataSource => insert(dataSource.id, dataSource))

Have a look at ReadWriteLock

[fm3]

The current dataSourceRepository will probably soon be fully replaced in #8384 so I’d say it’s not worth separately tackling that. We can close this issue then.

[fm3]

Solved by #8384 and #8791