There is a known issue with importing large datasets in Insight for Data Center in a short period of time, where the clustermessage table can fill up with information and cause performance issues.
A workaround - suggested by Atlassian - is to add a service that flushes the table every 12 hours, retaining the data for only 2880 minutes (or exactly 48 hours). This solution should prevent performance issues resulting from overloading of the clustermessage table.
Jira administrator permissions are required to complete this task:
- Click the cogwheel at far right to view the JIRA ADMINISTRATION options and select System.
- Scroll down to the ADVANCED section and select Services.
- Under Add Service, under Class, select Build-in services.
- Click on Cluster messaging flush service.
- Enter the following information:
- Name -
Cluster messaging flush service
- Class -
- Schedule -
0 0 4/12 * * ?
- Name -
- Click Add Service.
- Enter the following for Retention Period -
- Click Update.