Normally, the primary should be dedicated to core functions, as opposed to monitoring/remote monitoring tasks, so other than some local monitoring of the Primary and its resources, remote probes should NOT be deployed and running unless you must do it from the primary for some valid reason, e.g., the hub's location on that specific network allows you to reach the given target(s). In general, remote monitoring probes should be deployed and operate from a Secondary/remote hub.
baseline_engine
should be running on remote hubs, not the primary unless one of the probes deployed there has baselines enabled
prediction_engine
should be running on remote hubs, not the primary unlessyoure monitoring some IT asset from the Primary and using ToT and/or TTT.
audit (optional)
cm_data_import (optional)- discovery_agent (can be offloaded to another hub/secondary/remoite hub)
discovery_server (can be offloaded to a child robot connected to the Primary Hub)
especially IF the probe_discovery queue is frequently backing up and requires a significant increase in memory and you're close to the local machine's memory limit already)
qos_processor (optional unless needed for frequent origin changes)
remote monitoring probes such as net_connect, rsp, etc.
telemetry (optional)
custom probes (unless necessary to be run on the Primary)
You can choose to deactivate one or more of the probes listed above, or do the following to delete them from the Primary hub, but if you need to save the probe's configuration/files for any reason first take a backup of the probe directory and save it off in a folder.
Deactivate the probe
Using IM, Rt-Click and delete the probe
Delete the probe directory from the file system
Ensure that the probe is gone within IM
In later UIM versions, one or more probes may have been deprecated or no longer deployed as part of the UIM Primary hub install. Check with support if you have any questions on a particular probe and if it is required/necessary to be running on the Primary hub machine.