UIM FQDN for robots?

Document ID : KB000122720
Last Modified Date : 06/12/2018
Show Technical Document Details
Issue:
Internal need to convert short name to FQDN automatically for the robots.
Environment:
- UIM 8.5.1 or higher
Cause:
- Internal escalation process requirements
Resolution:
UIM cannot "force" discovery_server to always use the FQDN. This is because using the FQDN when one has been discovered, is the 'default' behavior; therefore, if a robot/machine is NOT showing up with FQDN it generally means we were unable to discover the FQDN in the first place.

Examine discovery_server.cfg - there is a key called /reconciliation/display_name_priority
you will see that the third entry is:

other.PrimaryDnsName.

Values include:

"other.RobotName,other.PrimaryDnsName,other.SysName,other.VMName,other.ComputerName,other.label,other.PrimaryIPV4Address,other.PrimaryIPV6Address,mon_config_service.PrimaryDnsName,mon_config_service.RobotName,mon_config_service.SysName,mon_config_service.VMName,mon_config_service.ComputerName,mon_config_service.label,mon_config_service.PrimaryIPV4Address,mon_config_service.PrimaryIPV6Address"

If you moved that to the front, and cold start the discovery_server, as soon as we have a FQDN for a device, that should be the name the device gets but if and only if the FQDN is available.

If not, discovery will look at whether the device has any of the 'other' attributes.

The first one (other.RobotName), ANY of the CM_DEVICE entries for the CM_COMPUTER_SYSTEM, will be used to name the entries.

Note that I believe if you make this change it would ONLY be applied to NEW devices but there is a callback on discvoery that you can use to trigger a reevaluation of all devices. Its called -> "reimport_devices_by_cs_ids”

Recorrelation is part of the official documentation and you can find it here:

https://docops.ca.com/ca-unified-infrastructure-management-probes/ga/en/alphabetical-probe-articles/discovery_server-discovery-server/device-correlation-troubleshooting

Its presented, at least for recorrelation/correlation troubleshooting but Im not sure if its a good way/safe way with no downsides for accomplishing this goal as we don't test it for the purpose  mentioned, but it is worth testing it out in a given TEST/DEV environment to be safe.