Devices stop responding after a number of connections

[hluhome]

Hi,
I've setup tinytuya first to control my heatpump. I was also using it to poll the state of the device, however, after a number of days the device stopped responding with error code 914. Only rebooting the heatpump helped. Since then I've changed the logic to interact minimally but it still seems to hang after a month or so. I reported this to the vendor.

However, since then I also set it up with 2 energy metering smart sockets - pulling the energy consumption every 2 minutes. Both sockets stopped working after 3 days (after being polled ~2200 times). So I wonder if there is something that causes these devices to fail - and what I can do to fix this...? Or even debug.... I only use the status() call to pull the data. And yes, everything seems fine from the Tuya Cloud.

[uzlonewolf]

Yes, this is a common problem, especially for v3.4 devices. #443 There are no known solutions at this time besides power cycling the device when it happens.

[jasonacox]

I sure wish we knew what is happening inside these devices that is causing that. There must be some memory stack or buffer overrun corrupting the execution code space. If we knew of a way to clear it, we could add that. I suspect Tuya never soak tests the local APIs for conditions like this. Or, possibly the app is sending something that is triggering the clear?

[hluhome]

Ok, so essentially this can be seen on any device using Local Tuya? Has this been reported to Tuya? Or is Tuya deprecating Local Tuya?

…

On Wed, Jan 15, 2025 at 7:25 AM Jason Cox ***@***.***> wrote: I sure wish we knew what is happening inside these devices that is causing that. There must be some memory stack or buffer overrun corrupting the execution code space. If we knew of a way to clear it, we could add that. I suspect Tuya never soak tests the local APIs for conditions like this. Or, possibly the app is sending something that is triggering the clear? — Reply to this email directly, view it on GitHub <#581 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BOO5OWKK2AZNLOPSMGVKZ332KXWNRAVCNFSM6AAAAABVEHEHG2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJRGY3TSNZVG4> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jasonacox]

No, I have 33 devices that do not do this. It is a small percentage that behave this way, but it seems to be more prominent with 3.4 devices.

[hluhome]

Hmm, the heatpump is a 3.4 device while the smart plugs are 3.3 devices.... So you have no workarounds or suggestions what I could do or test?

…

On Thu, Jan 16, 2025 at 8:53 AM Jason Cox ***@***.***> wrote: No, I have 33 devices that do not do this. It is a small percentage that behave this way, but it seems to be more prominent with 3.4 devices. — Reply to this email directly, view it on GitHub <#581 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BOO5OWNGTOMFAFADTJIUUJ32K5JOTAVCNFSM6AAAAABVEHEHG2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJUGY3DKOJQHA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jasonacox]

Are you seeing problems with the 3.3 devices, or just the heatpump?

If anyone has suggestion for these "stop responding" 3.4 devices, we would love to hear your ideas. My suggestion was to try the SmartLife app after the device locks up to see if it can unlock it. That may give us a clue as to what we could send to the device to free it.

[hluhome]

All the devices that are not working with local tuya are still working fine with the Cloud API. I have heard that some people have successfully monitored their devices using Localtuya/Home assistant. I will try that next.

…

On Fri, Jan 17, 2025 at 8:29 AM Jason Cox ***@***.***> wrote: Are you seeing problems with the 3.3 devices, or just the heatpump? If anyone has suggestion for these "stop responding" 3.4 devices, we would love to hear your ideas. My suggestion was to try the SmartLife app after the device locks up to see if it can unlock it. That may give us a clue as to what we could send to the device to free it. — Reply to this email directly, view it on GitHub <#581 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BOO5OWNHXGO324MBAS7I3FD2LCPMBAVCNFSM6AAAAABVEHEHG2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKOJXGUZDSMBXGA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[jasonacox]

Thanks @hluhome ! Let us know what you find out.

[uzlonewolf]

A few months ago I blocked multicast/broadcast from my IoT SSID (on the Unifi controller it's Settings -> Wireless Networks -> Edit -> "Block LAN to WLAN Multicast and Broadcast Data" checked). Ever since then, not one v3.4 device has dropped offline, though I am still losing 1 or 2 v3.3 devices per week.