@Riley_S 

All my smartplugs are already using static IPs, more accurately, static DHCP reservations in my router, which is a Comcast XB7.

The issue didn't only happen while the firmware was being updated. It also happened when the power went out and then came back on.

The APs take a while to come back online after the power outage, longer than it takes for the smartplugs to boot. All 50 smartplugs would have booted up simultaneously, and completed booting before all my APs were done booting up. About half the smartplugs waited for the APs to be fully up, and connected. The other half somehow gave up trying to connect complety.

IMO, this behavior from the smartplug is a bug. If the AP goes down, or is in the middle of booting up, or if the smartplug is done booting up before the AP is, the smartplug should not give up trying to connect to Wifi completely. It should keep trying until the AP comes online.

What's strange is that some of the smartplugs give up, and others don't give up. They are all on the same latest firmware, latest 1.08 version. Random bugs are hard to fix, but this needs a fix from TP-Link.

BTW, just did a very simple test :

1. disabled the SSID to which the smartplugs were connected on all 6 Unifi NanoHD APs

2. waited a minute or so

3. enabled that SSID on my XB7 gateway/router

Now, all the smartplugs that were previously connected to Unifi APs are in the solid orange state, and are no longer trying to connect to Wifi at all.

I would have expected them to migrate to the XB7, but no.

Looks like it's time to toggle breakers again. Sigh.

  @Riley_S

When the Kasa devices go offline, they show as being offline in the Kasa app, not under local control. I can't ping their static IP address.

If we were talking a straight incompatibility, none of the smartplugs would be connecting. But the behavior is that half of them went to this odd state.

As far as AP settings, someone on the Ubiquiti forum suggested I created a separate SSID for IOT devices, one which is on the 2.4 GHz band only, and with the default Wifi settings. I did so last thursday. It took me the better part of a day to move all 50 TP125 smartplugs to this new SSID.

I have not had a power outage since then, and have not tested the behavior. I could simulate one by turning off the main breaker, and may do so later today.

But from my point of view, there is a bug in the TP-Link firmware. I don't understand what the "solid orange" state signifies. If all the APs went down, all my other Wifi devices wait for APs to come up when booting. The TP125 should do the same and never go into the solid orange state.

  @Riley_S 

It takes much less than 10 minutes for the APs to come back up.

A normal reboot takes about 90 seconds.

For AP firmware updates, I have seen them be offline as long as 7 minutes (looking at smokeping graphs to determine this), but never 10 minutes. Usually, I don't update all 6 APs at once, I do it in a sequence. Here are the settings for my new IoT SSID, as listed in the Unifi controller :

Here are the settings for my main SSID on which I saw the problem after the outage last week :

As far as radio/channel settings, those are set per-AP. The channels are set to auto for the 2.4 GHz band. Width is set to 20 MHz.

Here is a copy of the settings for one of the APs (wired one) :

There are 2 wired APs and 4 wireless ones (meshed).

@Riley_S ,

FYI, I rebooted all my APs seconds apart earlier today as a test. All the KP125 smartplugs came back online properly.

It looks like the new settings on the separate SSID may have helped. The behavior I saw is still very strange, and I don't think it should ever happen. There is no good reason I can think of for the smartplug to give up after 10 minutes. Sometimes, the APs/network could be under maintenance for longer than that. If it's truly designed to stop trying after 10 minutes and require the user to manually unplug/cut the power to restart it, that is not a good design, IMO. But I didn't have >10 minutes of Wifi AP downtime in this case anyway.