That caused me some confusion since I just made a change of local only to two of the five KP115’s. I went to check on them and the app was locked up. All OK now
2 Likes
I have over 10 KS115 that are reporting off/NA in Sense and are online and/or powered on in the Kasa App if you want to check my account.
Just started this morning. Many are powered on as well.
Some are in local only mode on Sense. Others are not. That doesn’t appear to make a difference.
DHCP lease time is currently set to 5 minutes as I put the Sense/KASA network together. Reasoning: The KASA units originally took the first IP available and I then moved the MAC to where I wanted them long term. Did not want a long wait.
Note all KS300 still are working nominal. EDIT: Not after Sense Reboot.
Note Kitchen - Refrigerator is on Smartplug and showing off instead of NA
Roof Attic Fan is on as well yet showing off instead of NA.
Same for Garage - Freezer
All the NA/Off are showing up under Other now.
Previously showing properly.
Rebooted Sense to determine if that would make any difference.
Now 3 HSA300 exhibiting same issues.
It a took a reboot of my router for my zombie HS110’s to be reachable again. More and more, I’m honing in on the smart plugs interactions with my networking, as the bad actors.
As noted, HS300 are now affected.
Furthermore, I took 3 HS300 out and deleted. That should clear up 18 devices. Replaced with 3 KP115s.
The HS300 would not disappear from Sense App. Finally had to turn off TP-Link Connection in App. Of course, that killed all history.
Unfortunately, though the new KP115 were detected, none of my other KP115 are detected in Sense.
I know they are on network as they are all pulling IP Addresses. I can also see data being exchanged.
Nope. Router reboot didn’t bring them back for me.
Only showing the Sense detected devices, the 3 KP115 I added in last hour and strangely 1 KP115 added a week ago.
None of the other KP115 and HS300 detected.
5 days, 11 min later, all five KP115’s went to N/A again, INCLUDING the two that were on local only. So that makes no difference.
Every 5 days plus a few min, all five KP115’s simultaneously go N/A in the Sense app. Three remote controllable, two local only. All are still in the Kasa app, controllable (ON/OFF) and monitoring energy use.
Only an unplug/replug of each KP115, will bring each back into Sense.
I sincerely hope this is a typo…
I am still waiting on a response from TP-Link. I wish it was faster.
1 Like
I’m not sure how my UDM ascertains “UPTIME” for my Kasa smartplugs, but it looks like 4 of them at the top have experienced an outage since my last “master reset” about 8 days ago. Does anybody know Ubiquiti’s technique for assessing UPTIME ?
@kevin1 Uptime is clock time in seconds that the device is up and running. The Controller converts those seconds into H.M.S for ease of understanding. This is the time from the boot up stage and connection to the AP. If you unplug them then it should reset back to 0 after boot.
I have mentioned a few times that the plugs don’t disconnect from wifi. They remain connected and they continue to communicate with Kasa Cloud, respond to ARP requests and obtain new DHCP leases. In my case and it looks like in @Beachcomber case we also see sporadic connections/response to Sense. Others experience a complete loss to Sense. Regardless of the sporadic or complete failure, a common denominator is the continued presence on Kasa Cloud indicating a continual connection to the AP.
1 Like
I went in KASA app and deleted all the Kitchen Plugs, Garage Plugs and a few others.
Went through the tedious process off adding them back in. KASA knows they are there. Alexa knows they are there. I even get a pop up in Sense App a new TP115 has been found.
Yet, several units have yet to show up in Sense App.
I have tried two units 5 times.
No go.
Furthermore, several of the brand new KP115 I deployed last night to replace a HS300 and were working 12 hours ago have already gone NA in Sense App during that 12 hours.
Sense is just not seeing them. As 3rd party Alexa does, the common denominator is the Sense App.
I know a lot but certainly not a network engineer. You mind messaging me why this is bad while moving reserved IPs around? I do not feel like waiting hours or days to wait for the IPs to end up where I want them after install. As no MAC is listed on box or switch, I can only assign it after install when I look to find new MAC on Network.
This is not true. Alexa and many other providers use the cloud API which we have discussed in nauseum about the plugs remaining connected to the cloud. Sense is using the local connection, which is a totally different beast. I’m doing a lot of local tests with Hubitat and HA right now and I’m seeing similar issues to that of Sense.
5 Likes
Finding the MAC and setting a lease time of 5 minutes don’t have much to do with each other. Your devices will get a new IP if you reserved one for them just by doing a reboot/power pull. You don’t need to set lease times to 5 minutes.
I’ll PM you shortly.
This is brutal. There is clearly something special about your network.
At the moment, I’m out of things to try in your network other than running packet captures.
I’ve run a 10 day capture, but haven’t had time to explore the data yet.
Very good to know. I have mine enabled in Hubitat, but I don’t use them enough there to notice any issues. Still haven’t seen any issues with Sense.
By temporarily setting DHCP lease time at 5 minutes, I can get the MAC off the Desktop Screen, add it into the Router Reservation Table and have it update without needing to get up, go to to the plug, pull it from the outlet and then reconnect. Some of the plugs behind washer, freezer, refrigerators, pool pump, garage door motors and the like are not that convenient to get to anyway.
After the network is setup, I intended to move lease time back to a longer time period. Not that my Network is overtaxed as is.
At least that was my thought process. And I guess I just don’t understand why that’s a bad idea while building out the Network🤷🏼♂️
Thanks @DevOpsTodd, I was wondering more deeply how the Controller detects a disconnect, and whether there might be other things besides a smart plug lock-up or reboot that might look like a disconnect and consequent end/reset of UPTIME. As you suggest, I am seeing disconnects and reconnects that coincide with the start of a new UPTIME, though there are some cases where my controller reports a disconnect, but not the subsequent reconnect (Office HS110). Not sure if that is a reporting shortcoming or an artifact of some other issue.
Kevin,
It gets more complicated with each vendor. For example, with Cisco it will show a client offline when the data is below a threshold even though the client is actually online.
For UniFi devices I believe (in other words someone who might know better chime in) that they have no requirement for data. I cannot find anything online that would give indication that it’s based on data, ARP or some other metric.
But, that being said, just because a device is offline in a controller doesn’t mean it really is offline, at least with Cisco products.
1 Like
Thanks @DevOpsTodd, I posted the same question on the Ubiquiti community, but no answers yet. In the meantime, I’m watching my UPTIMES for the Kasa devices more carefully. It looks like I have about 5 of them that disconnect with greater frequency than the rest. It does look like my sole KP115 is the leader in disconnect/reconnect frequency. I probably should turn on more detailed logging for all of the Kasa devices, but that means going back to the Classic UI. Ugh.
