I’m confused. Seems that there are 3 different sets of WAN Priority settings in the configuration. They are:
WAN Connection Priority in the Dashboard settings
WAN Connection Priority in the SFC sub-tunnel settings
WAN/SFC Priority in the Outbound Policy settings when choosing Algorithm = Priority
Questions:
a) How do each of these work? For instance, it seems that I need multiple Priority 1’s in the first two cases above, but not really sure. Ranking 1 to 6 in the sub-tunnels doesn’t seem to get the results I’d expect.
b) How do they interact with each other? For instance, it seems that the Dashboard priority overrides the others.
Treat WAN priority as big clunky levers that open or shut a WAN link for use by everything else. It is responsible for WAN availability.
With no other configuration but this on a Peplink, user traffic will be load balanced all healthy WANs available in the highest priority. All other links will be unavilable.
If all WANs in P1 are unhealthy the WANs in P2 will be used. etc etc.
Dashboard WAN prioirty is usually used for least cost routing, the availability for each WAN in each priority is based on WAN health status (using whatever healthcheck is set on each one).
WAN priority in SpeedFusion is for SF traffic only. You can have a different set of priorities in the SpeedFusino config, but SF can only use the WAN links that are available to be used.
Outbound policy priority can be used to direct traffic over WANs or SFC tunnels in an order of preference. These are rules for application / user traffic flow more than about controlling the use of WAN links.
If you want to use all the WANs in a SpeedFusion tunnel they will need to be healthy and ‘available’ - so all in the currently live priority (normally P1).
Then you can decide to prefer one WAN over another using SpeedFusion priority. This then provides hot failover between those WANs.
As above. Dashboard for availability, Speedfusion for Speedfusion flow control, Outbound policy for application level control.
This is the most awesome & informative reply I have read for a while! Thank you.
Extra question: Where is the hierarchy of actions should I see the SpeedFusion Connect Route by LAN Client option? I assume it comes first and precedes (and maybe overrides?) any Outbound Policy rules? Similar for the other SpeedFusion Connect Route by Cloud App and Wi-Fi SSID?
Good question. When you create an Outbound Policy by default that policy affects traffic that has not already been managed by other routing protocols.
So if you have SpeedFusion VPN / OSPF / BGP / RIPv2 Routes or SpeedFusion Connect Routes & policies configured, those will be applied first to outbound traffic and then your outbound policies apply to the left over traffic that has not already been affected by those previous routing rules / policies.
You can however turn on expert mode in outbound policies and drag a rule above the ‘automatic routing protocols’ line if you want more control.
Thank you. And also for the hint that you have to turn on “expert mode” in the “?” popup for Outbound Policy to see this (didn’t see this previously).
Since I am on a roll in getting great answers: what is the relationship between setting up an SFC (and directing traffic to it with an Outbound Policy rule) and QoS Bandwidth Control? I don’t see how to use Bandwidth Control to ensure “reserved” bandwidth for the SFC tunnel in an environment where I also have non-SFC traffic (typically more uncontrolled and bursty random web stuff) going over the same WAN that is one of the WANs in the SFC? I would like to be able to classify the SFC use within my top QoS User Group so the tunnel gets priority in a crunch. Is this possible? Maybe use Subnet in some way?
