Hi all,

Having a weird issue, it may be related to my limited understanding of queues (so far, learning fast...). I have a hotspot setup on a bridge interface, and on the hotspot server profile, Rate Limit set to 18M/18M. Users are auth'd via RADIUS, and given a rate limit of 1M/1M. Thus, I see one dynamic queue created for each auth'd user, limited at 1M/1M, and a queue for the bridge interface (BR_HOTSPOT as target) limited to 18M/18M.

I also have PPPoE users on this interface, plus a bypassed IP range, so I thought if I add a simple queue with target BR_HOTSPOT, Max Limit 18M/18M, it would do a rate limit on the bridge interface no matter what the traffic source is.

If I enable the queue, the BR_HOTSPOT interface traffic graph jumps instantly to 18Mbps, even though I have 8-10Mbps of actual traffic without the queue enabled. If I disable the queue, traffic goes back to normal.

This is what I currently have (not adding the per-user dynamic queues to this):
My aim is to dedicate 18M/18M of the 20Mbps we have to all users on the BR_HOTSPOT interface, which includes hotspot users, PPPoE users, and devices on the bypassed IP range, so that we reserve 2M for "admin traffic" which includes things like analytics servers, web admin portals, etc.