I've got a Mesh point and Mesh portal (with AP-374s) both configured to a local (L2/LAN) controller with bridged mode AP wired Port profiles on them (not using Tunnel mode and are trusted) RSSI is fantastic and just around -30. These are being hosted by a local AOS 8.6 controller as Mesh Portal and Point respectively.
Both Mesh APs are working great and have minimal packet loss across to the Mesh Point which then serves a wired switch. That wired switch then hosts x4 CAPs that are up and operating fine, but some hosts utilizing a tunnel-mode virtual-ap off of those CAPs and are encountering IP packet loss while staying associated to the AP.
I've checked end to end small MTU issues and none exist. What could be causing some clients behind those CAPs via the Mesh topology to be getting packet loss while some (even associated to the same AP) encounter NO packet loss? This is less than 20 users total spread across a 4 CAP RF neighborhood. AP packet loss to the IPs of the CAPs themselves also encounters no loss. The MESH point and portal have AP Wired ports (both e0) running in bridged-mode as to avoid a tunnel-in-tunnel scenario with the tunneled virtual APs hosted by the CAPs via ap profiles running with tunnel-mode..
Looking for general best practices, lessons learned and general pitfalls for those of you that have deployed Aruba CAPs behind an Aruba Mesh Portal/Point and have dealt with tunnel-mode vs bridged-mode along with CAPs behind a Mesh Portal/Point and potential gotchas of this deployment topology.