[Q&A] Lan scan results in entities

[SirAry]

Documentation & bug reporting acknowledgment

Not applicable

Describe your problem

So there is a root company (entity) and a child one (let's say subsidiary or Partner, or whatever). The child entity SA wants to have his agents (*on his LAN) and inventory, but I don't want anything mixed with the root's ones, where I am the SA, as the LANs are different and each SA has responsibility only in his LAN. I want to know if the child entity agent deployed on it's LAN, may directly report the scanned result directly in it's entity and have that entity separated from root in this aspect.

[g-bougard]

You'll have to use GLPI rules to set entity based on some criteria. If you use a different ip numbering, you can use an ip criteria matching.

[SirAry]

This is a problem if as root SA I have to do rules for it. I do not have access at that LAN, it is another SA there, and as the policy of GLPI says that entities are used to entirely separate all this need breaks the separation, isn't it? Or the child entity SA may do those rules? But if he forgets to set a rule, and run some agent without, I will have all what the agent scans in root inventory and mess all there? It will be nice if at install to have a setting to directly specify to agent to what server and to what entity to send the result. Or am I not right?

[SirAry]

I see also that the agents appear by default in root entity and I do not see any possibility that an agent to be installed and appear in the entity where it is destinated to scan. Is it any way to achieve normality without modifying the agent, or this may be a feature request? It seems like a contradiction between the entity principle and the agent so I think that maybe somewhat I don't understand or it may be an unfinished work about glpi agent?