After the rack/setup/install tasks are completed, Serviceops needs a task to track the implementation.
This was done manually so far and is error prone, which risks that we don't use racks that are ready and waste resource.
We suggest adding an automation in the form of Herald rule on those tasks.
A production perspective of this is also to put an alert on cumin for insetup hosts: sudo cumin 'A:owner-serviceops and A:insetup' being run regularly would identify the stale hosts.