ACC-V and duplicate detection for non-IRE tables

[Jim Palmer]

I'm trying to manage discovery of operating system user accounts from multiple sources (including the ACC-V agent).

According to the documentation, the ACC-V Agent collects local user account data from the 'local_users.rb' ruby file and stores them persistently in the [cmdb_os_user] table in the instance.

My question is whether the ACC-V data for this table goes through the IRE or has a different method of managing duplicate detection? The OS Users table [cmdb_os_user] isn't a child of [cmdb_ci] so it doesn't turn up in the CI Class manager and it's not classified as a 'core' table so the 'IRE support for non-CMDB tables' doesn't apply either.

If you have the ACC agent managing duplication itself (not sure how because it just throws discovered data at the instance database) then how can you have another source also throw the same data at the instance and still manage updates/de-duplication when the target table [cmdb_os_user] are not included in IRE because they are not children of [cmdb_ci]?

This should also apply to other ACC-V populated entries that are not children of [cmdb_ci] like [cmdb_serial_number] and [cmdb_running_process].

[Jim Palmer · Accepted solution]

Never mind, found it:

There's a script include called "sn_acc_vis_content.EnhancedDiscoveryHandler" that takes the JSON payloads from the ACC agent for users, software, network adapters etc. (non-IRE classes) and runs glide record queries matching the JSON data to the instance records to execute gr.update() or gr.delete() actions.