Keeping DDNS updated with WAN changes

I have a total of two Balance 20s and 3 Surf SOHOs. (Soon to have a 210! :<) ) The issue I raise is common to all.

Generally, DDNS updating works quite well. However there is a circumstance where it does not

Suppose the router has two WAN connections:

  1.   Cable, 30 down / 5 up
    
  2.   DSL, 1.5 down / .5 up
    

Now suppose WAN1 fails. The DDNS will point to WAN2. WAN1 recovers, DDNS will still point to WAN2.

What I think is needed: A firmware change so the admin should be able to configure DDNS to be updated via the highest bandwidth WAN port available (or perhaps other criteria, if appropriate). When that WAN port fails, DDNS should update via another port, as selected by the admin. When the “main” or “preferred” port returns to service the DDNS should again update via that IP.

Here’s the rationale: Inbound traffic, addressed by the address name, say RICKROUTER.SCHMOOO.COM, should always go to the highest bandwidth port. If the highest capacity connection is restored and DDNS is not updated RICKROUTER.SCHMOO.COM will point to the slower port, in this case, a very slooow DSL connection. We don’t want to leave it in that situation. After all, that’s what Peplink/Pepwave does so well – fail-over and load balancing. What I propose would, I think, improve the performance of inbound connections when addressed by DDNS name. Certainly it would be “Pareto optimal” at the very least – no “down side” to those who do not want to take advantage of such a feature.

I know one Peplink owner who has an IP name for each of his WANs and his Balance keeps each updated nicely. Three ports and all three are frequently updated. But, I think this is a “workaround.” It would be better if the admin could configure a single DDNS name, and the router would keep it updated according to the preferences the admin has set. And, the preference that is presently missing , in my view, is "keep DDNS updated and assigned to highest capacity WAN port available.

Yes, exactly. There are two threads referencing the same issue. I am not the only one requesting this.