You are not logged in.
Hello,
I'm just starting to use GLPI and it is a wonderful project, thank you very much!
I have noticed that both the original XML inventory specification ( github.com/glpi-project/inventory_format/blob/master/original.md ) and the current JSON schema ( github.com/glpi-project/inventory_format/blob/master/inventory.schema.json ) include virtualmachine fields "image", "mac", "comment", "owner", and "serial", but these fields are not available in the interface and are also not in the "glpi_computervirtualmachines" database table. I am using the brand new GLPI release, 10.0.9. These fields are present in the XML generated by the latest FusionInventory client.
Am I correct that this support is currently missing in GLPI, both in the database and the frontend interface? Is this something that could be implemented, or is there any documentation on how to submit a PR to add fields like this?
Also, it seems like it would be very helpful if there was an option to have VM Images also listed in the Software inventory.
Thank You!
Offline
Hi jantman,
first, if you're starting to use GLPI, you should still use GLPI-Agent which is actively maintained where FusionInventory agent is not.
About you question, the XML spec is quite old and the JSON one is anyway based on it. Actually GLPI doesn't need all the spec to be supported. On the agent side, the virtualisation fields are not always retrieved, this will depend on the virtualisation technology being inventoried. Most of the time, you can also run the agent from the virtualmachine so it will complete the inventory. And this is the best to do to obtain a more accurate inventory. Have you tried this ?
If you want to develop something, there's official glpi documentation, just google for it or start from the official github project.
VM images are not software, why would this be very useful to list them there ?
Anyway can you also tell which virtualisation systems you're using as there are some and this can clarify your point of vue.
GLPI-Agent developer from Teclib' and GLPI-Network team
Previously FusionInventory-Agent maintainer
Offline
Hi jantman,
first, if you're starting to use GLPI, you should still use GLPI-Agent which is actively maintained where FusionInventory agent is not.
About you question, the XML spec is quite old and the JSON one is anyway based on it. Actually GLPI doesn't need all the spec to be supported. On the agent side, the virtualisation fields are not always retrieved, this will depend on the virtualisation technology being inventoried. Most of the time, you can also run the agent from the virtualmachine so it will complete the inventory. And this is the best to do to obtain a more accurate inventory. Have you tried this ?
If you want to develop something, there's official glpi documentation, just google for it or start from the official github project.
VM images are not software, why would this be very useful to list them there ?
Anyway can you also tell which virtualisation systems you're using as there are some and this can clarify your point of vue.
gbougard, thank you so much for your fast reply!
Ok, I will try glpi-agent, I had not even heard about that before. I had found fusioninventory-agent simple because it's already in the Debian repositories, I can just `apt install fusioninventory-agent`.
I know that the agent is collecting those fields, because I can see them in the XML... but I also don't see anywhere in GLPI for them to be stored.
Apologies, I probably should have mentioned that I'm only using Docker, which may explain why I'm asking about adding Docker images as software. From my point of view, Docker images *are* just software packages, and I need to try and keep them up to date with the latest release.
Offline
Now I understand your user point of view.
You should probably first ask on GLPI userecho site if this can make sens to have an option to inventory Docker images as software: https://glpi.userecho.com/
GLPI-Agent developer from Teclib' and GLPI-Network team
Previously FusionInventory-Agent maintainer
Offline