JVMPeg

Updated -

Introduction

JVMPeg was designed based on a popular solution we have for making analysis in situations where JVM threads are consuming a significant amount of CPU: Java application high CPU. When the JVM is consuming a lot of CPU we need the ability to peak inside the thread activity to see what specific thread is a high consumer. We've developed scripts on the aforementioned solution that will capture jstack and top output that can be used in JVMPeg for analysis.

For a command line version of analysis see javahighcpu

Usage

  • Download the appropriate high_cpu_*.tar.gz script from How do I identify high CPU utilization by Java threads on Linux/Solaris.
  • Run the script per instructions on the solution.
  • Drag and drop the resulting high-cpu.out and high-cpu-tdump.out onto the respective drop areas on the screen.
  • Optionally select the CPU threshold (default 40%). This will determine at what percentage of CPU consumption to begin showing threads.

Was this helpful?

We appreciate your feedback. Leave a comment if you would like to provide more detail.
It looks like we have some work to do. Leave a comment to let us know how we could improve.

Get notified when this content is updated

Follow

18 Comments

Subscriber exclusive content

An active Red Hat subscription is required to participate.

Log In

Great tool, makes the analysis much faster. Thanks !

Analysis does not start... Tried with 3 different browsers...! Is it really working?

Same here - accepts the dump files but never processes them, just keeps spinning ....

Firefox: Blank page, wont load the application at all Chrome: Allows files to be submitted - no reply from analysis Chromium: Same a Chrome

Hi Paul, we believe we identified an issue with the database backing this app. The database has been restarted and the app should be working again, please let us know if you have any further issues.

Hi Tomé, we believe we identified an issue with the database backing this app. The database has been restarted and the app should be working again, please let us know if you have any further issues.

Is this App working? Can not analize my files with any threshold.

Getting Could not parse high cpu top output, please use a file with valid input, please advise

Tool not working, tried so many times to parse my output. Better to use external thread dump analyzers for analysis, don't relay on this tool as it's not working with high_cpu_linux_jstack.sh script outputs

It seems that this tool does not work any more. I tested but just got a message "Could not parse high cpu top output, please use a file with valid input ". I try to test this tool by Chrome and FireFox. In both case, it fails with same message.

Getting the same error message here.

The problem is related to the date locale format. Change the local time configuration and timezone inside the script. Try something like this: export TZ="America/Toronto" export LC_TIME=en_US.utf8 The date print should look like this: Mon May 4 20:41:44 EDT 2020

The option to analyse remains greyed out, i.e. no option to run the report - after copying the out file into the drag-and-drop area. Could do with this tool working.

Hi Andrew,

Which .out file did you upload? You need to updoald high-cpu.out and high-cpu-tdump.out both to accurate drag-and drop zone.

And is the .out file you uploaded generated by the script from the bottom of How do I identify high CPU utilization by Java threads on Linux/Solaris?

Thanks.

Hi Thanks for the reply. I am uploading the high-cpu.out file as this is the only file that is generated. I use the high_cpu_linux.sh script to generate the out file. Thanks Andrew

Hi Andrew,

The app can not work without high-cpu-tdump.out.

The script should generate two .out files. You'd better check if there is something wrong or some error occurred when you run the script.

Thanks.

Hi

I have checked the usage of the high_cpu_linux.sh script, and it is fine. The second file is not generating. Are you sure that this script will generate both files? Or does it require the high_cpu_linux_jstack.sh script to be run?

Thanks

Hi Andrew,

You should run high_cpu_linux_jstack.sh. You don't need to run high_cpu_linux.sh. high_cpu_linux_jstack.sh will generate two files.

Thanks.

Hi

I think that is where the problem is. The document says you can use either script, but in fact it's the jstack script you have to use. And if you dont have jstack installed then the high_cpu_linux.sh script is the only option, and that script wont generate the two files required.

I think the document just needs to be a bit clearer.

Thanks