"ASYNC: Discovery - Sensors AWS Relations" jobs were sending a large number of Glide queries

[srikanthnelapud]

DESCRIPTION:

Whenever a Cloud or IP Discovery is made, then from either flow we try to find appropriate Guest (Eg. Windows Server, Linux Server, etc) / VM Instance (Eg. Record in cmdb_ci_vm_instance or cmdb_ci_vmware_instance table) to add the Virtualized by::Virtualizes relationship.

When the schedule is running in async mode i.e., at a regular time period based on the schedule frequency set, the Sensor(s) and Pattern Post Processing Script ends up calling the API - "findParentSysIdFromRelationsTable" of the script include - "DiscoveryAWSRelationshipSensor"

The queries made in the above API can return many hundreds of thousands of rows which can introduce significant latency thus hitting the performance and CPU usage.

RELATED ISSUE ADDRESSED VIA THE PROBLEM: PRB1479781 (Fixed in ROME)