Sometimes webservices seem to hang completely (T102102: Problems with multiple tools). Unfortunately, I did not take the time to strace the webservice to see what was going on at that point.

We should somehow check for webservices that hang. I'm not entirely sure how to do this, as we'd need to distinguish from 'this just took very long so nginx gave up'. Maybe add a standard /__check endpoint in the lighttpd config that just serves a static page...?