Java Performance Checklist

• Specify the required performance.
  • Ensure performance objectives are clear.
  • Specify target response times for as much of the system as possible.
  • Specify all variations in benchmarks, including expected response ranges (e.g., 80% of responses for X must fall within 3 seconds).
  • Include benchmarks for the full range of scaling expected (e.g., low to high numbers of users, data, files, file sizes, objects, etc.).
  • Specify and use a benchmark suite based on real user behavior. This is particularly important for multi-user benchmarks.
  • Agree on all target times with users, customers, managers, etc., before tuning.


• Make your benchmarks long enough: over five seconds is a good target.
  • Use elapsed time (wall-clock time) for the primary time measurements.
  • Ensure the benchmark harness does not interfere with the performance of the application.
  • Run benchmarks before starting tuning, and again after each tuning exercise.
  • Take care that you are not measuring artificial situations, such as full caches containing exactly the data needed for the test.


• Break down distributed application measurements into components, transfer layers, and network transfer times.

• Tune systematically: understand what affects the performance; define targets; tune; monitor and redefine targets when necessary.
  • Approach tuning scientifically: measure performance; identify bottlenecks; hypothesize on causes; test hypothesis; make changes; measure improved performance.
  • Determine which resources are limiting performance: CPU, memory, or I/O.
  • Accurately identify the causes of the performance problems before trying to tune them.
  • Use the strategy of identifying the main bottlenecks, fixing the easiest, then repeating.
  • Don't tune what does not need tuning. Avoid "fixing" non-bottlenecked parts of the application.
  • Measure that the tuning exercise has improved speed.
  • Target one bottleneck at a time. The application running characteristics can change after each alteration.
  • Improve a CPU limitation with faster code and better algorithms, and fewer short-lived objects.
  • Improve a system-memory limitation by using fewer objects or smaller long-lived objects.
  • Improve I/O limitations by targeted redesigns or speeding up I/O, perhaps by multi-threading the I/O.


• Work with user expectations to provide the appearance of better performance.
  • Hold back releasing tuning improvements until there is at least a 20% improvement in response times.
  • Avoid giving users a false expectation that a task will be finished sooner than it will.
  • Reduce the variation in response times. Bear in mind that users perceive the mean response time as the actual 90th percentile value of the response times.
  • Keep the user interface responsive at all times.
  • Aim to always give user feedback. The interface should not be dead for more than two seconds when carrying out tasks.
  • Provide the ability to abort or carry on alternative tasks.
  • Provide user-selectable tuning parameters where this makes sense.
  • Use threads to separate out potentially blocking functions.
  • Calculate "look-ahead" possibilities while the user response is awaited.
  • Provide partial data for viewing as soon as possible, without waiting for all requested data to be received.
  • Cache locally items that may be looked at again or recalculated.


• Quality-test the application after any optimizations have been made.


• Document optimizations fully in the code. Retain old code in comments.

Popular posts from this blog

Some Important Facts About Core Java

There are 4 important OOPS concepts in Java : Inheritance, Abstraction, Encapsulation and Polymorphism. There are two types of variables in Java : Primitive types and Object Types The default value of Boolean is false (the same applies for the Wrapper class too). Marker interface is the one which doesn't contains any methods , it is like empty interface. You cannot override the 'main' method of the Java, even-though if you try to do so it will not give any exception, program compiles and executes fine. The reason is : 'main' method is of type static . Static methods are class level methods. Even-though it looks like method overriding, it is not overriding. These are two different method with respect to the corresponding class. You can Overload the 'main' method of Java, it is valid in Java. You can have only one 'Public' class and any number of non-public classes in any java source file. Inheritance is an example for IS-A relationship. C...
Read more

How To Resolve "Java compiler level does not match the version of the installed Java project facet." Issue in Eclipse, STS tool or Eclipse Based Tools

Image
                  I have faced this issue during the Java 7 upgrade from Java 6 into one of our application. You will face this issue suppose if your application or project built on lower versions of Java (in my case it is Java 1.6) and you are trying to upgrade it to the Higher versions of Java (in My case it is Java 1.7), then Eclipse will throw this error in the tool. You will face this error even after changing the compiler version from build path. To fix this issue Follow the Below Mentioned Steps. 1.  Open the ' Markers ' Tab at the bottom of the eclipse or STS tool beside the ' Properties ' Tab. (as shown in the below figure) 2. Right Click on the Error as mentioned in the below figure. in The context menu click on the ' Quick Fix ' item as shown below in the figure. 3. Select the correct Java Project Facet from the ' Quick Fix ' Window. and click on the ' Finish ' Button. 4.  After Cli...
Read more

Resolve OutOfMemoryError With ExcelExport : Export Excel Utility with Apache POI Stream API (SXSSF)

                                 Whenever we try to export excel of huge data (for ex: around 200000-300000 records), most of the time we end up with OutOfMemoryError:JavaHeapSpace. And also consuming more time to process or export that much of data. Main reason to this kind of problem is that, the prior version of Apache POI (prior to 3.8) will not provide proper solution for this kind of situations and also we have other issues with the API designed with those versions. Even I had faced issues of not supporting more than 65000 rows of data during exporting excel with prior versions of POI. But with the version 3.8 and higher they come with solutions for all these problems. To resolve Memory issue and performance issue of Excel Export they have utilized stream API to design their API to support huge data export and perform...
Read more