We would like "LSN not changing" incidents to be given "error" status rather than "warning". Since late last year we experienced three incidents amounting to six connector outages where the above message appeared in the dashboard but no email notification was received. We think users should be informed immediately when their data is going stale rather than passively via the dashboard. I assume we did not receive an error notification by email because Fivetran treats this as a warning.
We also found the "Background info" in the warning message to be of little use and potentially (and unintentionally) misleading as, in each of our incidents, the relevant SQL Server Agent was running and did not require a restart.
Thanks for considering this!