You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As some of the devs know, I've been working on generating XML from Netdisco in order to leverage the large body of work there.
That is complete, BUT.... it is clear there are a number of definitions missing from the XML that are needed to improve FI4G's input to GLPI - this applies whether the source is FI-agent or ND or any other source.
One example: There is currently no way of telling GLPI which Vlans on a port are tagged and which are untagged, nor is there any way to link Alias ports (ip addresses) to aggregates or physical ones (It's best to do this, as IP addresses can "bounce" between ports, depending on what's active. Trying to conflate Layer2 and layer3 devices in one definition gets very complicated, very quickly..
Some discussion of the additions needed would be good - and for Vlans, we need to realise that it's perfectly possible to have several untagged vlans on an edge port, (the switch can separate them into their vlans based on MAC) and that PVID is not always the same thing as "untagged VLAN" - there's also the issue of QinQ - 4096*4096 vlans, not just 4k
The text was updated successfully, but these errors were encountered:
As some of the devs know, I've been working on generating XML from Netdisco in order to leverage the large body of work there.
That is complete, BUT.... it is clear there are a number of definitions missing from the XML that are needed to improve FI4G's input to GLPI - this applies whether the source is FI-agent or ND or any other source.
One example: There is currently no way of telling GLPI which Vlans on a port are tagged and which are untagged, nor is there any way to link Alias ports (ip addresses) to aggregates or physical ones (It's best to do this, as IP addresses can "bounce" between ports, depending on what's active. Trying to conflate Layer2 and layer3 devices in one definition gets very complicated, very quickly..
Some discussion of the additions needed would be good - and for Vlans, we need to realise that it's perfectly possible to have several untagged vlans on an edge port, (the switch can separate them into their vlans based on MAC) and that PVID is not always the same thing as "untagged VLAN" - there's also the issue of QinQ - 4096*4096 vlans, not just 4k
The text was updated successfully, but these errors were encountered: