Skip to end of banner
Go to start of banner

Whole Jira or BigPicture is slow

Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Next »

In case of contacting our Support Team via our portal or e-mail for the first time, please remember to:

  • Describe a problem in detail.

  • Collect the required data (described below) and attach it to the request.

  • Provide an exact timestamp when a problem occurred. 


If you have contacted the Atlassian Support Team regarding performance problems before and have a preliminary analysis, let us know and add our Support Team to that ticket. 

Problem:

You notice low performance throughout the whole Jira and/or BigPicture (high usage of server resources, e.g. high CPU usage) or the whole Jira system does not respond.

Required data to collect and send to the Support Team:

You need to generate thread dumps while you are experiencing performance problems. Collecting thread dumps when problems no longer exist is useless.


You can generate thread dumps a few times as comprehensive data can help our Support Team analyze the problem faster. 


Technical info

Gathering thread dumps

If your BigPicture website or Boxes are loading slowly and you would like our Support to diagnose this issue, it is recommended to generate a thread dump file. 

A thread dump file will provide essential information on the processes and threads currently running within the Java Virtual Machine and help our Support solve the issue faster. 

Follow the steps described below.

  1. Open your BigPicture in two different tabs (one of them will be needed to go to the technical settings and the other to reproduce the issue).

  2. From the first tab, navigate to BigPicture > App configuration > Advanced > Technical Info.

    • Go to the "Support data" section and click "Generate thread dumps". 

    • Set "Number of generated thread dumps" to "10" and "Sleep time between each thread dump generation" to "1". 

    • Click "Download" and immediately switch to another tab to reproduce the issue.

  3. From the second tab, reproduce the issue you have encountered before.

  4. Once you reproduce the issue, you need to wait until the file is downloaded.

  5. Now, you can attach the downloaded file to your Support ticket.

See the video for more information:

If Jira or BigPicture is unresponsive and you cannot access the App Configuration > Advanced > Technical info page to generate thread dumps, you can generate thread dumps using the commands:

For  jstack installed:

It is recommended to use jstack because thread dumps with CPU usage are in one file. 

  1. When you notice a performance problem, you can Identify the Jira application PID by using a command:

    JIRA_PID=`ps aux | grep -i jira | grep -i java | awk  -F '[ ]*' '{print $2}'`;


  2. Run the following command - it generates 10 snapshots of CPU usage and thread dumps with 10-second intervals for a minute. You must have a JDK installed for the jstack command to work.

    for i in $(seq 10); do top -b -H -p $JIRA_PID -n 1 > jira_cpu_usage.`date +%s`.txt; jstack -l $JIRA_PID > jira_threads.`date +%s`.txt; sleep 10; done


  3. Compress generated jira_cpu_usage.txt and jira_threads.txt files and send them to our Support Team.  

No jstack installed:

  1. If you do not have jstack installed, use the following command when you notice a performance problem:
    This command generates 10 snapshots of CPU usage and thread dumps with 10-second intervals.

    for i in $(seq 10); do top -b -H -p $JIRA_PID -n 1 > jira_cpu_usage.`date +%s`.txt; kill -3 $JIRA_PID; sleep 10; done

    When "kill -3" is used, the thread dump is sent to the standard error stream. If you are running your application in tomcat, the thread dump will be sent into <TOMCAT_HOME>/logs/catalina.out file, which is usually attached to the Jira support zip.

  2. Compress generated jira_cpu_usage.txt files and send them with the Jira support zip file to our Support Team.


You can also find instructions for all scripts recommended by Atlassian on this page


  • No labels