-
Notifications
You must be signed in to change notification settings - Fork 77
Automated Tests
The Eclipse performance test plugin (org.eclipse.test.performance) provides infrastructure for instrumenting programs to collect performance data and to assert that performance doesn't drop below a baseline. The infrastructure is supported on Windows, Linux, and MacOS X.
The first part of this document describes how performance tests are written and executed, the second part explains how performance data is collected in a database and how this database is installed and configured.
- check out the following plug-ins from
http://git.eclipse.org/c/platform/eclipse.platform.releng.git.
org.eclipse.test.performance
-
org.eclipse.test.performance.win32
(for Windows only)
- you need
org.junit
- add
org.eclipse.test.performance
to your test plug-in's dependencies
A performance test case is an ordinary JUnit test TestCase. Create a test case with test methods along the lines of:
public void testMyOperation() {
Performance perf= Performance.getDefault();
PerformanceMeter performanceMeter= perf.createPerformanceMeter(perf.getDefaultScenarioId(this));
try {
for (int i= 0; i < 10; i++) {
performanceMeter.start();
toMeasure();
performanceMeter.stop();
}
performanceMeter.commit();
perf.assertPerformance(performanceMeter);
} finally {
performanceMeter.dispose();
}
}
or create a test case extending PerformanceTestCase which is a convenience class that makes the use of PerformanceMeter transparent:
public class MyPerformanceTestCase extends PeformanceTestCase {
public void testMyOperation() {
for (int i= 0; i < 10; i++) {
startMeasuring();
toMeasure();
stopMeasuring();
}
commitMeasurements();
assertPerformance();
}
}
Notes:
- The scenario id passed to
createPerformanceMeter(...)
must be unique in a single test run and must be the same for each build. This enables comparisons between builds. ThePerformance#getDefaultScenarioId(...)
methods are provided for convenience. -
PerformanceMeter
supports repeated measurements by multiple invocations of thestart()
,stop()
sequence. The call tocommit()
is required before evaluation withassertPerformance()
anddispose()
is required before releasing the meter. - The first iterations of the above
for
-loop will generally take more time, because the code is not optimized by the JIT compiler yet. This can introduce some variance to the measurements, especially if other tests run before and change in some way that affects the JIT's optimization of the measured code. A simple way to stabilize the measurements is to run the code a few times before the measurements start. Caches also need special caution as they can affect the measurements. - As a rule of thumb the measured code should take at least 100 ms on the target machine in order for the measurements to be relevant. For example, Windows' and Linux 2.4's system time increases in 10 ms steps. In some cases the measured code can be invoked repeatedly to accumulate the elapsed time, however, it should be kept in mind that JIT could optimize this more aggressively than in real-world scenarios.
If the number of performance tests grows large, it becomes harder to get a good overview of the performance characteristics of a build. A solution for this problem is a performance summary chart that tries to condense a small subset of key performance tests into a chart that fits onto a single page. Currently the performance infrastructure supports two levels of summaries, one global and any number of "local" summaries. A local summary is typically associated with a component. A summary bar chart shows the performance development of about 20 tests relative to a reference build in an easy to grasp red/green presentation.
So dependent on the total number of components every Eclipse component can tag one or two tests for inclusion in a global and up to 20 for a local performance summary. Tests marked for the global summary are automatically included for a local summary.
Marking a test for inclusion is done by passing a performance meter into
the method Performance.tagAsGlobalSummary(...)
or
Performance.tagAsSummary(...)
. Both methods should be called outside
of start/stop calls, but it must be called before the the call to
commit()
.
// ....
Performance perf= Performance.getDefault();
PerformanceMeter pm= perf.createPerformanceMeter(perf.getDefaultScenarioId(this));
perf.tagAsGlobalSummary(pm, "A Short Name", Dimension.CPU_TIME);
try {
// ...
In order to keep the overview graph small, only a single dimension
(CPU_TIME
, USED_JAVA_HEAP
etc.) of the test's data is shown and only
a short name is used to label the data (instead of the rather long
scenario ID). Both the short label as well as the dimension must be
supplied in the calls to tagAsGlobalSummary
and tagAsSummary
. The
available dimensions can be found in
org.eclipse.test.performance.Dimension
.
The PerformanceTestCase
provides similar methods that must be called
before startMeasuring()
:
public class MyPerformanceTestCase extends PerformanceTestCase {
public void testMyOperation() {
tagAsSummary("A Short Name", Dimension.CPU_TIME);
for (int i= 0; i < 10; i++) {
startMeasuring();
toMeasure();
stopMeasuring();
}
commitMeasurements();
assertPerformance();
}
}
- Create a new JUnit Plug-in Test launch configuration for the test case.
- Adjust -Xms and -Xmx if needed. Note that during automated production runs, the settings are set to what the eclipse.ini has. If your test needs some different (larger or smaller) be sure to set both (ms and mx) in your test.xml file, using the vmargs property.
- Run the launch configuration.
- By default, the measured averages of the performance monitor are
written to the console on
commit()
. This is suppressed if performance tests are configured to store data in the database (see below).
- This method is of particular interest if you want to do precise measurements of a specific build.
- Export the following plug-ins as 'Deployable plug-ins' to the
eclipse installation directory (not the 'plugins' directory; choose
to deploy as 'a directory structure')
org.eclipse.test
org.eclipse.test.performance
-
org.eclipse.test.performance.win32
(for Windows only) - your test plug-in(s)
Copy the following .bat file and customize it to your needs (Windows)
@echo off
REM Installation paths
SET ECLIPSEPATH=c:\eclipse
SET JVMPATH=c:\jdk\jdk1.4.2_05
REM Paths, relative to ECLIPSEPATH
SET BUILD=I200411050810
SET WORKSPACE=workspace\performance
REM Test
SET TESTPLUGIN=org.eclipse.jdt.text.tests
SET TESTCLASS=org.eclipse.jdt.text.tests.performance.OpenQuickOutlineTest
REM For headless tests use: org.eclipse.test.coretestapplication
SET APPLICATION=org.eclipse.test.uitestapplication
REM Add -clean when the installation changes
SET OPTIONS=-console -consolelog -showlocation
SET JVMOPTIONS=-Xms256M -Xmx256M
ECHO Build: %ECLIPSEPATH%\%BUILD%
ECHO Workspace: %ECLIPSEPATH%\%WORKSPACE%
ECHO Test: %TESTPLUGIN%\%TESTCLASS%
%JVMPATH%\bin\java %JVMOPTIONS% -cp %ECLIPSEPATH%\%BUILD%\startup.jar org.eclipse.core.launcher.Main %OPTIONS% -application %APPLICATION% -data %ECLIPSEPATH%\%WORKSPACE% -testPluginName %TESTPLUGIN% -className %TESTCLASS%
pause
After testing the setup you may want to close other applications to avoid distortion of the measurements.
If the test.xml
of your test plug-in already exists and looks similar
to the jdt.text.tests
' one, add targets similar to those shown below.
The performance target is the entry point for performance testing like
the run target is for correctness testing.
<!-- This target defines the performance tests that need to be run. -->
<target name="performance-suite">
<property name="your-performance-folder" value="${eclipse-home}/your_performance_folder"/>
<delete dir="${your-performance-folder}" quiet="true"/>
<ant target="ui-test" antfile="${library-file}" dir="${eclipse-home}">
<property name="data-dir" value="${your-performance-folder}"/>
<property name="plugin-name" value="${plugin-name}"/>
<property name="classname" value="<your fully qualified test case class name>"/>
</ant>
</target>
<!-- This target runs the performance test suite. Any actions that need to happen -->
<!-- after all the tests have been run should go here. -->
<target name="performance" depends="init,performance-suite,cleanup">
<ant target="collect" antfile="${library-file}" dir="${eclipse-home}">
<property name="includes" value="org*.xml"/>
<property name="output-file" value="${plugin-name}.xml"/>
</ant>
</target>
Notes:
- Do not include an empty performance target, in your test.xml file. Committers sometimes do as a reminder of where it would go, if/when they have performance tests, but some tools in the Maven/Tycho build (eclipse-cbi-plugin) produce a 'test.properties' file that describes the test bundles and characteristics, so an empty performance target causes a lot of noise and confusion (if not, later, processing complications). ()
- Performance measurements are run on (primarily) a dedicated performance measurement machine of the Releng team.
- If you have created a new source folder, do not forget to include it
in the build (add it to the
build.properties
file).
- Modify
test.xml
as described above. - Download a build and its accompanying test framework plug-ins
contained in
eclipse-test-framework-*.zip
- Unzip the Eclipse SDK and the eclipse-test-framework ZIP to install your target Eclipse.
(you need Info-ZIP UnZip version 5.41 or later (Windows) installed and added to the path).
- Export your test plug-in as a 'Deployable plug-in' to the target Eclipse installed above; choose to export as directory structure.
- Open a terminal or command window and execute the following (on a single line):
java -jar
/startup.jar
-application org.eclipse.ant.core.antRunner
-file
<target eclipse install/plugins/your test plug-in id_version>/test.xml
performance
-Dos=
-Dws=
-Darch=
-Declipse_home=
"-Dvmargs=-Xms256M -Xmx256M"
-logger org.apache.tools.ant.DefaultLogger
- The JUnit results are written to an XML file in the root of the target Eclipse and the performance measurements are written to the console.
- Performance (see especially for its section with links to related pages)]]
- Platform UI/Performance Tests (see for some early work on the framework, including a link to an EclipseCon2005 presentation!)
- Another 2005 writeup.
- A good overview tutorial by Lars Vogel.