TP-Link Plugin v 0.9.3

[jay (support)]

I had to quit using Kasa devices - for me, they fall off the network every couple of months and require power cycling to get them back (unplug, wait a while, then plug them back in). I think their WiFi implementation is sketchy...

Yeah, basically I have the same problem Jay has

@jtburgess

Any chance you will be updating the plugin for Indigo 2022.1 with Python 3 support on macOS 12.3?

I get the following errors in the log and assuming it is a Python issue? This is on an HS110 which does support Energy Monitoring.

Code: Select all

TP-Link Devices Error           (tpl_relay_poll.py:run:322): Error attempting to update new device: 'power_mw'. Will try again in 2 seconds
TP-Link Devices Error           (tpl_relay_poll.py:run:317): Unable to update new device: after 10 attempts. Polling for this device will now shut down. ('power_mw')

I have updated Indigo to 2022.1 (but have not yet updated MacOS to 12.3) and my TP-Link Kasa devices are working normally.

Following up to add -- I upgraded to MacOs 12.3 and Indigo 2022.1 and this plugin is working fine.

Just curious if you have devices with energy monitoring?

And do you have polling turned on?

Hs300 power strips. Yes polling on. And I have the issue wheee the drop from wifi and get disabled but that predates 2022.1.

I found that if I turn Polling on and Energy monitor off, everything is fine.

As soon as I turn Energy monitor on, it gives the errors and disables the plugin.

I also tried going back a few versions all the way to 0.9.3, but still can't get the plugin to work...

Maybe the issue is unique to the HS110? Can you ping it reliably? These devices are known for poor wifi.

If it helps here's a sample log entry:

Code: Select all

TP-Link Devices                 (tpl_relay_poll.py:run:78): TV Room Apple TV connection received ({"system":{"get_sysinfo":{"sw_ver":"1.0.21 Build 210524 Rel.161309","hw_ver":"1.0","model":"HS300(US)","deviceId":"REDACTED","oemId":"REDACTED","hwId":"REDACTED","rssi":-62,"longitude_i":-REDACTED,"latitude_i": REDACTED,"alias":"TP-LINK_Power Strip_78C1","status":"new","mic_type":"IOT.SMARTPLUGSWITCH","feature":"TIM:ENE","mac":"REDACTED","updating":0,"led_off":1,"children":[{"id":"REDACTED","state":1,"alias":"TV Room Camera HDMI","on_time":156364,"next_action":{"type":-1}},{"id":"REDACTED","state":1,"alias":"TV Room Dash","on_time":70754,"next_action":{"type":-1}},{"id":"REDACTED","state":1,"alias":"TV Room Apple TV","on_time":3836365,"next_action":{"type":-1}},{"id":"REDACTED","state":1,"alias":"TV Room Yamaha","on_time":156377,"next_action":{"type":-1}},{"id":"REDACTED","state":1,"alias":"TV Room Sonos","on_time":70732,"next_action":{"type":-1}},{"id":"REDACTED","state":1,"alias":"TV Room Mini","on_time":3836365,"next_action":{"type":-1}}],"child_num":6,"err_code":0}}})

I have been polling these HS110 devices for years with a Python Script to get real-time energy usage, and I have never had them drop (I have a Unifi AP very close to them). I have maybe 3 years of Grafana data, with no hiccups due to Wi-Fi.

Problem is that the reliable script I was using is Python 2, and since I am not a developer, I am not sure how to convert it to Python 3. Jay said in another thread it would be non-trivial.

Turns out the developer of the original script ported it to Python 3, but I can't quite figure out how to make it work, since they are a lot of improvements and additions to the script.

In any case, I am hoping the original developer of this plugin is still alive and willing to take a look.

Sumocomputers: As we discussed, your device has a different set of attributes (e.g., power, voltage). I now recognize those, without dividing by 1000.

Ramais: It recognizes the original attributes (e.g., power_ma, voltage_mv) as mill values, and divides the result by 1000 before saving to normalize as the original code did.

This is available on GitHub as 0.9.8.7. I will create an indigo tag once the next topic is resolved.

All : I can change it so that it never disables polling, but just slows down the polling rate and report lack of response every 5 tries.
It will resume normal polling when the device does eventually respond. What say you? Would this be a better behavior than arbitrarily disabling the device, which I too have seen and agree is annoying.

I like the polling changes. Instead of 5 for reporting interval, can it be made configurable? Reporting interval and time interval (and if you want to increase the time by X for eachre try, that would be cool too).

I am seeing the same thing as others where it will work fine for a couple of months and then randomly the communication is disabled in indigo. at first, I thought it was a Network issue since their address would change on a power cycle so I gave them a fixed IP address in the Unifi switch and that seemed to keep everything happy after a power cycle, Now I am seeing usually 1 or 2 will randomly have their communications check box get magically unchecked. re-enabling communications makes them work again. I do not have any triggers, schedules, or scripts that are disabling communication. Any ideas or is this a bug?

Mac OS: High Sierra
Indigo: 2021.2
PLugin Ver: 0.9.8

I am seeing the same thing as others where it will work fine for a couple of months and then randomly the communication is disabled in indigo. at first, I thought it was a Network issue since their address would change on a power cycle so I gave them a fixed IP address in the Unifi switch and that seemed to keep everything happy after a power cycle, Now I am seeing usually 1 or 2 will randomly have their communications check box get magically unchecked. re-enabling communications makes them work again. I do not have any triggers, schedules, or scripts that are disabling communication. Any ideas or is this a bug?

Mac OS: High Sierra
Indigo: 2021.2
PLugin Ver: 0.9.8

Grab the latest version from GitHub:

https://github.com/jtburgess/indigo-TP-Link

I think it’s version 0.9.8.7