Stoiko,
To answer your questions in reverse order. Yes, the manually running of the command fixed the disk usage issue.
We had been receiving spam reports, but it appears they were only coming from the other node in cluster not this one. I assume we should we receive notifications directly...