A number of contributors have added some useful monitoring check list to supplement this good article. Great work everyone!

One thing I would be interested in finding out: what happens next? For example, if the monitoring shows up excessive disk queueing, how can this be fixed? Are there articles/books that discuss both monitoring and resolutions to the problem?