Being newbie in web-development I was really confused when I've don't seen any log entries while reloading demodomain.com page. But I figured out "buffer=32k" trait and "access_log off" directive for static files. And now I wonder: it is generally good or bad idea (for production site) - do not log static files requests? Can I meansure traffic for specified virtual host without it? Maybe it's more useful to process log files with awstats or similar tool? Awstats can aggregate summary old info in files like awstats052014.txt and then you can start new log for new day. Which approach do you guys use commonly?