How do we set Bandwidth limitations for users going/using the SpeedFusion route? For some reason, users who are set to use the SpeedFusion connection set in outbound policy are exceeding their bandwidth limitations.
Thanks for the links. I have read the first one before and checked out the other two just now. Some of the scenarios would apply but in my case having the HQ WAN 1 bonded with the remote WAN 1 doesn’t fall through the same crack. In your opinion, what would be the realistic speed we should be getting with that setup?
We are planning to replace our HQ connections with two high-speed fiber lines and have both of them bonded together to take advantage of the feature but if the gain is marginal then we might have to pull it off. Any suggestions, please? Thanks once again. Forum is very helpful
In practice, there are loads of things that can affect the amount of bandwidth available over a speefusion link.
Latency is the big one. Since SpeedFusion is a point to point link what we often see is raw speedtest.net measurements showing near line speeds (at low latency) but when going point to point between VPN endpoints - particularly when traffic goes between/across ISPs, latency can rise considerably. The higher the latency the lower the throughput.
DSL circuits can also be heavily contended, ISPs can do some crazy things to VPN traffic (they sometimes categorise it as bittorrent traffic so throttle it) I’ve even see specific routing paths (like those that cross inter country / continent boundaries) get saturated. So deep analysis is needed of the traffic flows and routing paths.
There are a bunch of ways to mitigate latency variation and packet loss on WAN links (which both affect speedfusion throughput), but fine tuning SpeedFusion can be an involved process - that’s why @sitloongs has suggested you get your Peplink partner involved.