Performance issues come with typical symptoms, such as response times, that are slow or become slower over time, timeouts, or even completely unavailable services. The error rates indicate the latter.

When performance issues arise, the question to be asked is what the actual constraining resource, the bottleneck, is. Where does the issue originate? As shown earlier, engineers are advised to follow an investigative process that considers the overall situation, including hardware and operating systems, in order to find the constraint. There should be no guessing and premature decisions.

Performance problems can have a huge number of root causes. Most of them originate in coding errors or misconfiguration rather than actual workload exceeding the available resources. Modern application servers can handle a lot of load until the performance becomes an issue.

However, experience shows that there are typical performance issue root causes. The following will show you the most serious ones.

Engineers are instructed to investigate issues properly, without following supposedly best practices and premature optimizations.