Wouldn't that require you to actively watch the logs going by?
lol, in that case, yes. we had 500 servers reporting centralized syslog and many people would just tail -f the central log to make sure not too much crazy shit was happening.
(in a perfect software system (spherical cow) obviously all errors would be tagged and categorized with proper monitoring and alerting thresholds. but, in the real world you have a 800,000 lines of php spread across 1,000 files written by 300 mostly junior people (who only stay for 6 to 14 months at a time) over the past 10 years. you make the best of what you've got.)
lol, in that case, yes. we had 500 servers reporting centralized syslog and many people would just tail -f the central log to make sure not too much crazy shit was happening.
(in a perfect software system (spherical cow) obviously all errors would be tagged and categorized with proper monitoring and alerting thresholds. but, in the real world you have a 800,000 lines of php spread across 1,000 files written by 300 mostly junior people (who only stay for 6 to 14 months at a time) over the past 10 years. you make the best of what you've got.)