Yet, what admin has an hour daily to ensure "due care"? In real life, the admins will check the servers only if something appears to be wrong with them. In an ideal world, the admins should be notified every time a errors or warnings are recorded in the server logs. Various monitoring solutions are available on the market, some quite complex, but many are trying to do too much or are reporting the wrong things. A PDF file with pie charts showing the distribution of events per server is pretty much useless. The cost of such solution may also become an issue even for bigger companies and add yet another burden to the administrators' shoulders.