Probably a stupid question. I have several aps connected with dynamic wds. At one point in time I added access lists to not allow some ap’s to connect to each other. I did it because I though they shouldn’t be connected if they have bad signal to each other. My question is If they aren’t connected to each other then are they then not part of the wds and there fore just creating interference for each other? Should I just let them all connect and then the paths will be figured out by the MAC layer?
We have large WDS network also, Rather than use access lists to prevent unwanted associations with dynamic WDS, we use static WDS and only set up connections that make sense.
We specify a default bridge on the WDS setup so that each AP will automatically add its WDS link as a bridge port when the WDS link comes up. The port cost will be set based on the link quality, so the best links are used to form the mesh.
You could use dynamic WDS for this and just let the poor performing links hang around, unused. They won’t be used because the port costs will be high due the poor link performance.
The overhead of WDS is low if you don’t have any traffic going over the link.
However, your log file will be filled with link down/link up messages if the link is really marginal. This is what led us to use static WDS to avoid trying to set up links that will just flail.
Your idea of using access lists makes sense. Would that be easier, though, than using static WDS and only establishing links that you really want?