Due to live restore workers opening a cursor on a whatever collection they are migrating, enabling or disabling logging causes the mongod to crash.

Changing the state of logging on a given collection requires exclusive access, the background thread intentionally prevents this behaviour. Now for restore purposes the flag wiredTigerSkipTableLoggingChecksOnStartup prevents this behaviour but I can foresee on prem customers and developers running into this crash scenario.

The solution here, I think is to log a good error message and still crash or bubble up an error code to the caller or the log enable code path.

I don't know how practical it will be to check the relevant live restore state.