Peformance Issues

1) I’m running out of memory, and the Validatar service crashes, and/or CPU is running at 100%.

Answer:
This could be due to various reasons. Check the following:

• Test Configurations:
  • Review and optimize test result configurations.
  • We recommend using an Order By clause in your queries and changing the Scripts Include Order By setting to Yes in result configurations.
• For Python Data Sources:
  • Ensure efficient script design.
    • Avoid overusing for loops; consider set logic for better efficiency.
    • Use streaming when processing files.
  • Be cautious about overusing Pandas for file processing.
  • Consider pushing processing to the cloud rather than downloading large files.
  • Install a data agent on a separate dedicated server to run resource-intensive Python scripts.

2) Validatar site isn’t loading.

Answer:
Check the following:

• Ensure that the application pool is running, and restart it if needed.
• Verify domain account credentials, ensuring the password is not expired, and the correct account is used. To update credentials that have recently changed:
  • Inside of IIS Manager, click on Application Pools, then right-click the Validatar pool and choose Advanced Settings from the popup menu. Within that dialog, click on the Identity row in the Process Model section and then click the "..." button at the right. Click the Set button on the popup and enter the new credentials. For domain accounts, the username should use the following format: <domain><username>.
• Check if the account used for running the application pool is disabled.

3) My job isn’t running.

Answer:
Investigate the following:

• Check for server and database server reboots.
• If the job is scheduled to run when IIS is configured to restart, it may impact job execution.
• Scheduled jobs won't kick off if the job is still running from a previously scheduled run. Consider aborting the previous run and reviewing schedule timing versus how long it takes the job to execute.

4) My database size is increasing.

Answer:
Consider the following:

• Review test result configurations.
• We recommend turning on settings such as Abort Processing After X Failures, Purging Results, and Only Keeping Failures.
• Explore switching to the Powered by Snowflake Data Processing Engine, which saves to the Snowflake Repository for more efficient data storage.

If you have further concerns or specific issues, feel free to reach out to our support team for assistance.