Hi Fabrice,
thanks for your help. I forgot to mention that I have already tried turning on multihost, which did not change the behaviour. I turned it back of though because I did not notice any change and didn't know what this setting changes in the background. As I mentioned, the funny thing is that the issue I experience only arrises when I have multiple devices on a port that have different roles assigned. Benjamin ________________________________ Von: Durand fabrice via PacketFence-users <[email protected]> Gesendet: Donnerstag, 29. August 2019 01:26:07 An: [email protected] Cc: Durand fabrice Betreff: Re: [PacketFence-users] Multiple Devices on one Switchport Hello Benjamin, it looks that this is what you need https://github.com/inverse-inc/packetfence/pull/2034 it's in PacketFence since the version 7.0, can you try to enable advanced.multihost and retry ? Regards Fabrice Le 19-08-28 à 08 h 25, Shirley, Benjamin via PacketFence-users a écrit : Hello everyone, from reading up I am aware my setup is not officially supported, BUT it works on our equipment with one caveat I would love to get some information on. We have a lot of offices equipped with workgroup switches connected to (Aruba) HP 2920-48G access switches. The access switch is able to provide multiple vlans per port and separation of clients does work. For example I can have 1 client on the workgroup switch being able to access the registration vlan and another client accessing production network, all controlled by mac authentication and the radius-reply by Packetfence. The problem is that in the situations where I have multiple vlans on that one access switch port, Packetfence closes locationlog entries for some of the nodes and no more switch port information is available, thus reevaluating access or restarting of switch ports does not work when changing role or deleting a node. If all clients on the workgroup share the same role / vlan I can see the switchport details and the before said access re-evalution works / the switchport restarts when changing node role. (off course all clients loose connectivity for the moment, but this is acceptable) I would like to understand which procedure triggers this behavior and if there is any chance that I can get this working? Thanks in advance Benjamin _______________________________________________ PacketFence-users mailing list [email protected]<mailto:[email protected]> https://lists.sourceforge.net/lists/listinfo/packetfence-users
_______________________________________________ PacketFence-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/packetfence-users
