Performance troubleshooting - info Clarity Support needs

Products

• Clarity PPM SaaS
• Clarity PPM On Premise

Issue/Introduction

If you are seeing perfomance issues, there are some details that Clarity support will request to assist in troubleshooting.

One or more of the following symptoms are reported:

• Slowness across all pages in the Clarity application.
• You receive a blank page upon login.
• All actions require a wait of up to a minute before the action occurs.
• Jaspersoft reporting is slow.
• The performance is poor after a migration or upgrade.
• Portlet Export To Excel gives Error 500 - Internal Server Error. The server could not retrieve the document due to server configuration or other technical problems.
• Jobs that are in scheduled status never go to the processing state.
• A significant increase in job runtimes by 30-50%.

Environment

Release: All Supported Releases

Cause

The problem can occur at any point in the network, ranging from the application to the end server.

Resolution

When opening a case with the Support team to review why Clarity is slow, provide the following:

1. What is the exact issue with performance
2. When did the problem start to occur?
3. Is the issue intermittent or can it be reproduced at will?
4. How many users are affected?
   • If only one user is reporting the issue, what action is the user taking to reproduce the issue?
   • Are they accessing the URL remotely or within the office network?
   • Has a browser trace been created for the affected user? (This will show the time it takes to load page(s))
   • If many users are affected, which region is affected?
5. Does a user with all administrative rights experience the issue?
6. Go to the Administration > Security and Diagnostics > Log Analysis page:
   • Check if there is any issue coinciding with what the userbase is reporting compared to the generated graphs.In order for the graphs to be populated, make sure the job (Tomcat access log import/analyze) is run on a daily basis.
   • If there was any recent increase in activity around exporting to excel (particularly large amounts of data), XOG, or processes, this can cause slowness.
7. Has the issue been reviewed by your network team?
   • Has a trace been created and analyzed by the network team? With the assistance of your network team, a network trace can be generated for analysis. The packet trace can then be analyzed by the network team to pinpoint any potential bottlenecks or dropped packets.
   • Is the local network aware of any outage or bottleneck happening at a particular node or switch?
8. What page/portlets is the slow performance seen on?
9. If the issue is with a particular job, a runtime history may be requested by Support.
10. Is there high memory or I/O usage on the application server? (On Premise only)
11. Has the database been checked for issues? (On Premise only)
    • If it is a suspected database issue, obtain and attach the following to the case in a ZIP/RAR file:
      • For Oracle, provide the html version of the AWR report.
      • For SQL Server, provide the Top Performance reports.
      • For PostgreSQL, provide the Badger reports 
    • The DBA team can review inactive/blocking/active sessions that are long-running and terminate/kill the sessions.
    • Additionally, the DBA team can verify the DB server health check and check for abnormal memory or swap usage.
    • Make sure the database requirements from the documentation have not changed or there have not been any network/server changes.

Additional Information

For Broadcom-hosted GCP/SaaS environments, if there is a platform stack issue, notifications will be sent out to all customers affected,
however an alert will not get sent out at a tenant level.