Wyze Cam v3 – cameras occasionally crash during day/night mode changes

[adrian58001]

Hi all,
I've been running 8 Wyze v3 cameras with Thingino fw and have been chasing a few stability issues for a while. I've made quite a few changes and things have improved quite a bit, almost to the point of being stable but I've still got one problem I haven't been able to get to the bottom of.

I'm not even sure whether this is a Thingino issue, a camera hardware issue or something in my network setup, so thought I'd ask here before jumping to any conclusions.

My setup:
8x Wyze Cam v3 cameras running Thingino.
6 are outdoor cameras and 2 indoor cameras.
UCTRONICS UC-3AF-USB Ethernet-01 PoE-to-Ethernet adapters.
TP-Link TL-SG1210MPE 8-port PoE+ switch (123 W budget).
Cameras are on a dedicated IoT VLAN.
Frigate 0.18 running on 32GB ram N100 TrueNAS Scale in Dockge with a USB Coral.
All cameras running main stream for record and sub stream for detect. Main stream 15fps 2000bitrate CBR 15 GOP 1920x1080 audio enabled AAC, sub stream 5fps 600 bitrate CBR 5 GOP 768x432 audio disabled.
Home Assistant handles the MQTT commands for changing camera day/night modes using sunrise/sunset. I disabled photosensing on boot.
All cameras are T31X / GC2053 / RTL8189FTV

I previously had quite a few general stability problems, particularly when running all the cameras together. Since making some changes, including adding tmp/cache, things have improved a lot. I can now run the cameras much more reliably, so I'm not seeing the same level of general instability I was experiencing earlier.

The problem I'm still having is with day/night mode changes.
The outdoor cameras are controlled by Home Assistant through MQTT. I've also staggered the mode changes with a 3-minute delay between each camera to avoid having several cameras change modes at the same time. I disabled photosensing on boot in all the cameras.

Despite this, an outdoor camera changing from day to night (or vice versa) will sometimes cause all the cameras to crash or become unreachable. It doesn't happen every time, though. Some days the whole transition runs without any problems, while on other days the cameras drop out during the process.

The two indoor cameras are handled separately. Home Assistant uses their brightness sensors to trigger the day/night mode changes through an automation. So far, I've been using the same approach of allowing time between changes rather than sending commands to multiple cameras together.

What I can't work out is why a mode change on one outdoor camera can occasionally bring everything down, even with the commands staggered by three minutes.

A few things I've already looked at:
The failures have occurred with Frigate stopped, so I don't think Frigate is the cause.
I've tested different numbers of cameras connected.
I've tried different unmanaged POE switch with no change.

The cameras use USB Ethernet adapters based on the RTL8152 chipset and cdc_ether driver.
I've checked interface statistics and haven't seen anything particularly obvious in the way of RX/TX or CRC errors.
I do get this in syslog though when the cameras fail before becoming unreachable - - :RtspServer.cpp]: RESPONSE CSeq=1358 status=200 len=31
2026-10-28 10:09:18,536 from 192.168.52.103: <12>Oct 28 21:09:18 prudynt[802]: [WARN:RtspServer.cpp]: ch0 RTSP send queue 864911B -- client falling behind, 32 frames dropped in last 5s

At present I am relying on the TP-Link switch POE auto recovery rebooting the cameras when the problem happens.
I'm wondering if anyone has seen similar behaviour when changing day/night modes, particularly with multiple cameras using Ethernet adapters.

Could the mode change itself be causing something to hang or crash in the camera? Is there anything specific in the logs I should be looking for to distinguish a camera crash from a network disconnect or power issue?

I'm not trying to suggest this is necessarily a Thingino problem. I'm just trying to narrow down where the issue is, as I've spent a fair bit of time chasing what I thought were network or power problems and don't want to head down the wrong path.

Happy to provide logs, configuration or any other details that might help.

Thanks.
Adrian

[virmaior]

My setup is rather different (USB NCM), but I've also seen some instability at the daynight switch moment. I think there can be a big power draw right at the moment the camera switches day/night due to the IR lights coming on and the filter being moved.

[adrian58001]

Yeah I have suspected the same, I might see if I can cobble together a breakout lead for the usb cable and see if I can replicate it on the bench with my oscilloscope