How to read a heapdump generated in .phd format

Technote (troubleshooting)


Problem(Abstract)

Starting with Java™ SDK 1.4.2 SR1, the default heapdump format has changed. The default format for a heapdump will be "portable heap dump format" and the file will have an "*.phd" extension.

Cause

The default heapdump format has changed starting with Java SDK 1.4.2 SR1.

Resolving the problem

Starting with Java SDK 1.4.2 SR1, the default heapdump format has changed. The default format for a heapdump will be portable heap dump format and the file will have an *.phd extension. As most of the time to produce a heapdump file is spent in writing the file to the disk, it is very likely that by using the PHD format we will be able to make the entire process faster. This was made possible by using a Java DataOutputStream class which allows the output of Java primitive data types in a portable binary format.


PHD format heapdumps have the following benefits:

· Includes tracking information of all objects that move during compaction
· Includes information of all global references
· Includes the time and number of the GC cycle corresponding to the heapdump.
· Faster generation of the headumps thus performance benefits.
· Easy to upload and the file itself is in a compressed format compared to normal .txt format

How do I process and analyse the heapdumps generated in .phd format

Please use svcdump.jar attached in the Related Information section. This version of svcdump.jar works with heaproots HR205 version.

Note: Heap roots HR207 version can read *.phd format heapdumps

How to use svcdump.jar with heaproots

Download and copy the svcdump.jar in the same location where you have HR205.jar and then use the following sample command:

C:\heap> java -Xbootclasspath/p:svcdump.jar -Xmx768M -jar HR205.jar -i heapdump.20041105.181252.25035.phd

To obtain a copy of heaproots please see Related Information at bottom of this technote.


Enabling HeapDumps in .txt format for WebSphere Application Server V5.1.1 (any fix level) releases

  1. To set the following Environment Entries from the administrative console, select:
    Servers > Application Servers > server_name> Process Definition > Environment Entries > New
  2. Add the following Name and Value pairs:
    Name
    Value
    Description
    IBM_JAVA_HEAPDUMP_TEXT
    true
    Setting is variable will enable the headump to be generated in .txt format. Default is .phd format
    IBM_JAVA_HEAPDUMP_TEST
    true
    Setting is variable will enable the headumps generated in .txt and .phd formats
  3. Verify that changes are saved to the master configuration
  4. Restart the Application Server

Related information

Download latest svcdump.jar

Cross reference information
Segment Product Component Platform Version Edition
Application Servers WebSphere Application Server Enterprise
Business Integration WebSphere Business Integration Server Foundation
Application Servers WebSphere Application Server - Express
Application Servers Runtimes for Java Technology Java SDK

Rate this page:

(0 users)Average rating

Add comments

Document information


More support for:

WebSphere Application Server
Out of Memory

Software version:

6.0.2, 6.1, 7.0, 8.0

Operating system(s):

AIX, HP-UX, Linux, Solaris, Windows

Software edition:

Base, Enterprise, Express, Network Deployment, WebSphere Business Integrations Server Foundation

Reference #:

1190476

Modified date:

2004-11-16

Translate my page

Machine Translation

Content navigation