Fae wrote

So the ER605 router is connected behind the old Cisco SPA122 ATA router, and its WAN connection is set with "Dynamic IP", right?

 

If possible, you may visit the controller and check the CPU/Memory of the ER605 router to see whether there are abnormal changes before the WAN drops. In addition, check the Log and find the related logs for analysis.

  @Fae Thanks for replying.  The WAN connection is set with "Static IP" and not "Dynamic IP" because my ISP issued me a static IP and this is also how it was set in the Apple Time Capsule configuration.

I've checked the controller charts to see only Traffic and Packets went to 0 on the last outage, but no visible oddities about memory and CPU.  The only logs I can find are focused on the controller, not the ER605, and only flag up the issues related to consequences of the controller losing access to the cloud connection.  Are there ER605-specific logs somewhere in my setup that might contain root causes of the outages?

I really appreciate your helping me diagnose this.

Regards

  @watou This error held off for several days, but returned: connection stops working with nothing in the logs and only a reboot of the ER605 will fix it.  It may be 3 days or more before it happens again, with packets not moving.  A reboot of the router fixes the problem every time, but that's not sustainable.

Are they any other users using a Static IP route that have this problem?  Any suggested workarounds or new firmware releases on the horizon?  Thank you.

  @KimcheeGUN Thank you very much for your suggestion.  Much appreciated.

My ISP decided to put the SPA in front, and since it's their equipment, I have just gone along with it because for years it never showed any issue when my router was the old Apple Time Capsule sitting in the same position and with the same static IP/netmask/default route.

  @KimcheeGUN I know the SPA has a public IP and then routes to my 605 and its public IP.  Since the SPA is ISP-owned equipment, I doubt that they want to put it on my side of my router.  But you've made me think that maybe that old SPA may just not be a great router if something about the packets has changed when I switched from the Time Capsule to the 605, and rebooting the 605 only has the side-effect of clearing the SPA's problem?  In other words it takes two to tango, and maybe the 605 has only surfaced an issue in the SPA?

My obstacle now is that the ISP would have to come out and re-plumb the WAN connection somehow, since I have no admin access beyond my router.  I've been very happy with my ISP, but if fiber comes around soon from a competitor, this will all be a moot issue.

Fae wrote

How about your network size? Are there any applications/servers with high flow requirements running in your network, making the sessions filled and resulting in the connection stop working?

It's a small home+office network with meager Internet traffic, and the WAN failures have happened around roughly 4am or when no one is actively using the network.  It doesn't seem coincident with any particular usage pattern.  If anything were logged it would be helpful, but when it fails all one can see is 0% Gateway usage on the Omada dashboard instead of the typical 1.2% (varies) usage, and no packets moving to/from the WAN.

Thanks for your thoughts and attention!