This was reported against the System Health->Data Aggregator Queries Dashboard in Performance Management. The errors were observed in the "OpenAPI Query Processing Times" and "RIB Query Processing Times" Views. The Dashboard was set to use the 'All Groups' Group for Context.
This may be observed for other Views and Dashboards.
The top bar below the View title will show something like "1 of 2 queries failed. See logs for details."
Expanding the error drop down reveals the following error in the web UI:
Node3 in the 3 node DR cluster had it's /tmp space full. This prevented the package upgrades from completing post upgrade after a DB restart. The old package is what triggered the error.
Performance Management r3.6
Cleared out /tmp on Node3 and restarted the DB.
RIB Debug: https://comm.support.ca.com/kb/how-to-debug-report-problems-in-ca-performance-management/kb000101388