blob: 6639e725063be8690e1d016f9ea72d62ade99ea8 [file] [log] [blame]
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE task PUBLIC "-//OASIS//DTD DITA Task//EN" "task.dtd" >
<!--
Copyright (c) 2008, 2023 SAP AG, IBM Corporation and others.
All rights reserved. This program and the accompanying materials
are made available under the terms of the Eclipse Public License 2.0
which accompanies this distribution, and is available at
https://www.eclipse.org/legal/epl-2.0/
SPDX-License-Identifier: EPL-2.0
Contributors:
SAP AG - initial API and implementation
IBM Corporation - minor updates
-->
<task id="task_basictutorial" xml:lang="en-us">
<title>Basic Tutorial</title>
<prolog>
<copyright>
<copyryear year=""></copyryear>
<copyrholder>
Copyright (c) 2008, 2023 SAP AG, IBM Corporation and others.
All rights reserved. This program and the accompanying materials
are made available under the terms of the Eclipse Public License 2.0
which accompanies this distribution, and is available at
https://www.eclipse.org/legal/epl-2.0/
</copyrholder>
</copyright>
</prolog>
<taskbody>
<prereq>
<p>
If you are using Memory Analyzer installed into Eclipse rather than
a stand-alone Memory Analyzer,
first open the 'Memory Analysis'
perspective using:
<menucascade>
<uicontrol>Window</uicontrol>
<uicontrol>Perspective</uicontrol>
<uicontrol>Open Perspective</uicontrol>
<uicontrol>Other ...</uicontrol>
<uicontrol>Memory Analysis</uicontrol>
</menucascade>
</p>
</prereq>
<context>
<p>
This tutorial provides a "jumping-off place" to get familiar with
Memory Analyzer.
</p>
</context>
<steps>
<step>
<cmd>
<b>Get a Heap Dump</b>
</cmd>
<info>
<p>
The Memory Analyzer works with
<xref href="../concepts/heapdump.dita">heap dumps</xref>
. Such a heap dump contains information about all Java objects
alive
at a given point in time. All current Java Virtual Machines
can
write heap dumps, but the exact steps depend on vendor, version
and
operation system. Find out more in the section
<xref href="../tasks/acquiringheapdump.dita" />
.
</p>
</info>
<substeps>
<substep>
<cmd>
Open
<image href="../mimes/me76fc4.png">
<alt></alt>
</image>
<xref format="html" scope="peer"
href='javascript:liveAction(%22org.eclipse.mat.ui%22,%22org.eclipse.mat.ui.snapshot.actions.OpenSampleHeapDumpAction%22,%22org.eclipse.mat.ui.help/heapdump/HeapDumpSample.hprof%22)'>
a sample heap dump
</xref>
if you view this page inside the Eclipse help center.
</cmd>
</substep>
<substep>
<cmd>Start your application with Java 6</cmd>
<info>
<p>
For the purpose of this tutorial, we use Java 6 and JConsole
on
Windows.
</p>
</info>
</substep>
<substep>
<cmd>
Start
<codeph>&lt;jre6&gt;/bin/jconsole.exe</codeph>
and select the running application (in this case Eclipse):
</cmd>
<stepresult>
<image href="basictutorial_jconsole_open.png">
<alt>JConsole dialog to open a connection to a Virtual Machine.
</alt>
</image>
</stepresult>
</substep>
<substep>
<cmd>
Select the operation
<i>dumpHeap</i>
from the
<i>com.sun.management.HotSpotDiagnostic</i>
MBean.
</cmd>
<info>
The first parameter
<i>p0</i>
is the full path to the heap dump file. Make sure you give it the
file extension .hprof. The second parameter
<i>p1</i>
should be left at true as we are only interested in live objects.
</info>
<stepresult>
<image href="basictutorial_jconsole_mbean.png">
<alt>Select the dumpHeap method of the HotspotDiagnostics mbean.
</alt>
</image>
</stepresult>
</substep>
<substep>
<cmd>
Select the operation
<i>triggerDumpToFile</i>
from the
<i>openj9.lang.management=OpenJ9Diagnostics</i>
MBean.
</cmd>
<info>
<ul>
<li>For OpenJ9/Semeru based JVMs (build openj9-0.18.0 and later), the parameters are:
<parml>
<plentry>
<pt>p0</pt>
<pd>
is the dump type and should be one of <userinput>java</userinput>, <userinput>heap</userinput> or <userinput>system</userinput>.
</pd>
</plentry>
<plentry>
<pt>p1</pt>
<pd>
is the full path to the dump file or directory. Make sure you give it the
appropriate file extension <userinput>.txt</userinput>, <userinput>.phd</userinput> or <userinput>.dmp</userinput>.
</pd>
</plentry>
</parml>
See <xref format="html" scope="external" href="https://eclipse.dev/openj9/docs/api/jdk17/jdk.management/openj9/lang/management/OpenJ9DiagnosticsMXBean.html">API documentation for <codeph>Interface OpenJ9DiagnosticsMXBean</codeph></xref>
</li>
</ul>
</info>
<stepresult>
Note the name of the file.
</stepresult>
</substep>
<substep>
<cmd>Acquire a heap dump directly using Eclipse Memory Analyzer
</cmd>
<info>
<p>
See
<xref
href="../tasks/acquiringheapdump.dita#task_acquiringheapdump/acquire">
Acquire Heap Dump from Memory Analyzer
</xref>
to acquire a heap dump from a locally running Java process
directly using Eclipse Memory Analyzer.
The actual Eclipse Memory
Analyzer process could be used as a
example Java process to
generate a example heap dump.
</p>
</info>
</substep>
</substeps>
</step>
<step id="overview">
<cmd>
<b>Examine the Overview</b>
</cmd>
<info>
Open the heap dump via
<menucascade>
<uicontrol>File</uicontrol>
<uicontrol>
<image href="../mimes/i-openhd.png">
<alt></alt>
</image>
Open Heap Dump...
</uicontrol>
</menucascade>
to see the overview page.
If you have a heap dump available, try <xref format="html" scope="peer" href="javascript:executeCommand(%22org.eclipse.mat.ui.actions.openHeapDump%22)">
<b>Open Heap Dump</b></xref> now.
</info>
<stepresult>
<image href="basictutorial_overview.png">
<alt>Memory Analyzer's overview page for a heap dump</alt>
</image>
<p>
On the right, you'll find the size of the dump and the number of
classes, objects and class loaders.
</p>
<p>
If the total size of the dump is much smaller than the size of the
file it is possible
that the heap dump contained many 'garbage'
objects which would be
discarded at the next garbage
collection. See
the
<xref
href="../reference/inspections/unreachable_objects.dita">unreachable objects</xref>
query to examine these 'garbage' objects.
</p>
<p>
Right below, the pie chart gives an impression on the biggest
objects in the dump. Move your mouse over a slice to see the
details
of the objects in the object inspector on the left. Click
on any
slice to drill down and follow for example the outgoing
references.
</p>
</stepresult>
</step>
<step id="histogram">
<cmd>
<b>Get the Histogram</b>
</cmd>
<info>
Select the
<xref format="html" scope="peer" href="javascript:executeCommand(%22org.eclipse.mat.ui.actions.executeInspection(org.eclipse.mat.ui.actions.executeInspection.commandName=histogram)%22)">
<uicontrol>
<image href="../mimes/i-histogram.png">
<alt>histogram icon</alt>
</image>
histogram
</uicontrol>
icon
</xref>
from the tool bar to list the number of instances per class,
the
<xref href="../concepts/shallowretainedheap.dita">shallow size</xref>
and the
<xref href="../concepts/shallowretainedheap.dita">retained size</xref>
.
</info>
<stepresult>
<image href="basictutorial_histogram.png">
<alt>Histogram</alt>
</image>
<p>
The Memory Analyzer displays by default the retained size of
individual objects. However, the retained size of a set of objects
-
in this case all instances of a particular class - needs to be
calculated.
</p>
<p>
To approximate the retained sizes for all rows, pick
<image href="../mimes/i-calcrs.png">
<alt>Calculate retained size</alt>
</image>
icon
from the tool bar. Alternatively, select a couple rows and use
the
context menu.
</p>
<image href="basictutorial_calc_retained.png">
<alt>Select calculate retained sizes from the tool bar</alt>
</image>
<p>
Using the
<b>context menu</b>
, you can drill-down into the set of objects
which the selected row
represents. For example, you can list the
objects with outgoing or
incoming references. Or group the objects
by the value of an
attribute. Or group the collections by their
size. Or or or...
</p>
<p>
One thing that makes the Memory Analyzer so powerful is the fact
that one can run any action on any set of objects. Just drill down
and slice your objects the way you need them.
</p>
<image href="basictutorial_context_menu.png">
<alt>Drill down via the context menu</alt>
</image>
<p>
Another important feature is the facility to
<b>group any histogram by class loader, packages or superclass</b>
.
</p>
<image href="basictutorial_group_by.png">
<alt>Group the histogram by class loader or package via the tool
bar
</alt>
</image>
<p>
Any decent application loads different components by different
class loaders. The Memory Analyzer attaches a meaningful label to
the class loader - in the case of OSGi bundles it is the bundle
id.
Therefore it becomes a lot easier to divide the heap dump into
smaller parts.
</p>
<p>
More:
<xref href="../tasks/analyzingclassloader.dita" />
</p>
<image href="basictutorail_by_classloader.png">
<alt>Histogram grouped by class loader</alt>
</image>
<p>
Grouping the histogram by packages allows to drill-down along
the
Java package hierarchy.
</p>
<image href="basictutorail_by_package.png">
<alt>Histogram grouped by packages</alt>
</image>
<p>
Grouping the histogram by superclass provides an easy way to
find for
example all the subclasses of java.util.AbstractMap,
etc...
</p>
<image href="basictutorial_by_superclass.png">
<alt>Histogram grouped by superclass</alt>
</image>
</stepresult>
</step>
<step>
<cmd>
<b>View the Dominator Tree</b>
</cmd>
<info>
<p>
Select the
<xref format="html" scope="peer" href="javascript:executeCommand(%22org.eclipse.mat.ui.actions.executeInspection(org.eclipse.mat.ui.actions.executeInspection.commandName=dominator_tree)%22)">
<uicontrol>
<image href="../mimes/i-dominatortree.png">
<alt>dominator tree icon</alt>
</image>
dominator tree
</uicontrol>
icon
</xref>
from the tool bar to display the dominator tree.
The
<xref href="../concepts/dominatortree.dita">dominator tree</xref>
displays the biggest objects in the heap dump. The next level of
the
tree lists those objects that would be garbage collected if all
incoming references to the parent node were removed.
</p>
<p>
The dominator tree is a powerful tool to investigate which
objects
keep which other objects alive. Again, the tree can be
grouped by
class loader (e.g. components) and packages to ease the
analysis.
</p>
</info>
<stepresult>
<image href="basictutorial_dominator_tree.png">
<alt>Dominator Tree</alt>
</image>
</stepresult>
</step>
<step>
<cmd>
<b>Path to GC Roots</b>
</cmd>
<info>
<p>
<xref href="../concepts/gcroots.dita">Garbage Collections Roots (GC roots)</xref>
are objects that are kept alive by the Virtual Machines itself.
These include for example the thread objects of the threads
currently running, objects currently on the call stack and classes
loaded by the system class loader.
</p>
<p>
The (reverse) reference chain from an object to a GC root - the
so
called path to GC roots - explains why the object cannot be
garbage
collected. The path helps solving the classical memory leak
in
Java:
those leaks exist because an object is still referenced
even though
the program logic will not access the object anymore.
</p>
<image href="basictutorial_path_menu.png">
<alt>Select path to GC roots from the context menu</alt>
</image>
</info>
<stepresult>
<p>
Initially, the GC root reached by the shortest path is selected.
</p>
<image href="basictutorial_path.png">
<alt>Path to GC roots</alt>
</image>
</stepresult>
</step>
<step>
<cmd>
<b>Generate the Leak Report</b>
</cmd>
<info>
<p>
The Memory Analyzer can inspect the heap dump for leak suspects,
e.g. objects or set of objects which are suspiciously big.
</p>
<image href="basictutorial_run_leak_suspects.png">
<alt>Run the leak report</alt>
</image>
<p>
Learn more in this blog posting:
<xref format="html" scope="external"
href="https://memoryanalyzer.blogspot.com/2008/05/automated-heap-dump-analysis-finding.html">Automated Heap Dump Analysis: Finding Memory Leaks with One
Click
</xref>
.
</p>
</info>
<stepresult>
See
<xref href="../tasks/runningleaksuspectreport.dita">running leak suspect report</xref>
for information about the report.
</stepresult>
</step>
</steps>
</taskbody>
<related-links>
<link href="../concepts/dominatortree.dita" />
<link href="../concepts/gcroots.dita" />
</related-links>
</task>