Zigbee2mqtt: No matching payload found for entity: Ikea Plug_update_available with state topic: zigbee2mqtt/Ikea Plug. Payload: , with value template Template("{{ value_json.update_available}}")

[TASSDevon]

Bug Report

Getting quite a few of these messages for my ikea plugs:
No matching payload found for entity: Ikea Plug_update_available with state topic: zigbee2mqtt/Ikea Plug. Payload: , with value template Template("{{ value_json.update_available}}")

How to reproduce it (minimal and precise)

I just updated z2m to the latest version

Debug Info

zigbee2mqtt version: 1.11.0

[Koenkk]

The update_available property is available after the device checked for an OTA update once. This will probably happen in 24 hours from now. You can force it by executing the check yourself: https://www.zigbee2mqtt.io/information/ota_updates.html#check-if-an-update-is-available

[CodeFinder2]

@Koenkk Would there be a way to avoid such warnings completely? I am also getting a lot of them at startup now and have seen them in the past for an Aqara Water Leak sensor as already mentioned here. After the update to v1.12.0 (or maybe even already to v1.11.0 -- can't remember), there are quite a lot of them spamming my logs. 🙈

EDIT: maybe update_available should be "initialized" to false at start (although no check is performed) so it can still be set to true once the first check is executed and an update is actually found later? This way, HA would not complain (or at least not SO often?)... ?

[Koenkk]

@CodeFinder2 good idea, added in the latest dev branch, let me know if it works.

[CodeFinder2]

@Koenkk great, thanks! Would it be possible to also assume false for the Xiaomi Aqara Water Leak sensor (for the same reason, see #1015 ) until a real value was provided, i.e., to have a default (false) there as well? Would be great!

I just paired¹ another water leak sensor and Z2M reports (via MQTT):

zigbee2mqtt:info  2020-03-26 21:37:51: MQTT publish: topic 'zigbee2mqtt/basement/ground', payload '{"battery":100,"voltage":3055,"linkquality":138,"last_seen":"2020-03-26T20:37:51.932Z"}'

As you can see, no "water_leak" attribute is present. I think it's totaly safe to assume false unless the sensor provides an actual value.

1: Pairing Xiaomi devices is soooo great in Z2M now, reallly cool! Great work! :-)

[Koenkk]

@CodeFinder2 good idea, implemented in the latest dev branch, can you check if it works?

[CodeFinder2]

Will definitely test ASAP when I find a free time slot - sorry for the delay!
EDIT: Any way to use zigbee2mqtt-edge HassIO addon to test this? I guess no, since it currently not updated / reflecting the dev branch, correct?

[Koenkk]

It is, just remove and install edge addon.

[CodeFinder2]

Ah good to know and it should be safe to:

• stop the non-edge addon version
• copy the entire config dir to a new dir (e.g., z2m-edge/)
• start z2m-edge addon with that new config dir and the exact same config like the non-edge addon
• after testing, stop z2m-edge and start z2m addon for normal operation again?

So, no changes in PAN ID or network key...safe to go this way? Don't wanna mess up my setup :see_no_evil:

[Koenkk]

yes!

[CodeFinder2]

Hm, any way to check if it's really using your latest dev commits? It says:

info  2020-03-28 00:00:36: Starting zigbee2mqtt version 1.12.0-dev (commit #38cb519d3cdeb9503620f47e6a81e250a16e0765)

And that commit hash refers to z2m addon commit 9 days ago...

EDIT: and...:

abo@titan:~/Development/zigbee2mqtt$ git lg
* ac7ac1a - (HEAD -> dev, origin/dev) Update converters. (vor 3 Stunden) <Koen Kanters>
* d73bcf3 - Support Sylvania iQBR30 (#3205) (vor 6 Stunden) <presslab-us>

EDIT2: maybe related to https://github.com/danielwelch/hassio-zigbee2mqtt/issues/242 and https://github.com/danielwelch/hassio-zigbee2mqtt/issues/284

[Koenkk]

@CodeFinder2 it was not indeed not updating anymore, fixed now, new build should be up in 4 hours from now.

[CodeFinder2]

Thank you so much, will test when it's online!

Yeah, I had this feeling that it might not be updated anymore ... :see_no_evil:

[CodeFinder2]

@Koenkk nonetheless, how can I check if its using your latest dev branch?

[CodeFinder2]

@CodeFinder2 good idea, implemented in the latest dev branch, can you check if it works?

TL;DR: all fine!
Details:

info  2020-03-29 00:32:14: Starting zigbee2mqtt version 1.12.0-dev (commit #52d4193bc714a52a52e30801d2ab15d07695886b)
[...]
info  2020-03-29 00:32:50: MQTT publish: topic 'zigbee2mqtt/basement/ground', payload '{"battery":100,"voltage":3025,"linkquality":138,"last_seen":"2020-03-28T23:32:50.677Z","water_leak":false}'

And this sensor has never seen any water yet, so, according to my understanding, it must be due to your fix! Plus: it doesn't show the "water_leak":false if I am going back to v1.12.0. :sunglasses:

My HA logs are clean (regarding the update_available and the water_leak payload warnings) as well now (also tested with v1.12.0 vs latest dev).

Again: great work, very appreciated! Thank you!

[Koenkk]

Thanks, ill close this