The network is unstable...is it Tapo's cloud?
Hi, I have 4x Tapo C320WS cameras installed and 1x C200. When trying to access these cameras via the TAPO app I frequently get "the network is unstable", and often a failure to connect. This happens both when I am using the app from my local network and when I access the app while not on my local network.
Each of these cameras uses a MicroSD card instead of the TAPO cloud. The SD cards are a mix of Sandisk SDSQUAB-128G-GN6MT and 1x Amazon Basics 128GB. They are all Class 10 MicroSD cards.
The "network is unstable" messages occur when viewing both live streams of the camera and when viewing videos stored on the SD cards.
There is no predictable pattern to when I will get "the network unstable" messages. Sometimes I will get one and the camera will proceed fine, with the exception that changing speeds on recordings won't work, and even then only for a minute. Sometimes all cameras will give this error and refuse to allow live or recorded viewing for up to half an hour.
During the periods where I get these messages the cameras continue to perform their duty of recording my cats. Push notifications for those motion events end up delayed during the "network is unstable" message periods.
I have two I tenets service providers:
1) Telus, providing a 2-line bonded VDSL connection providing a rock-solid 30Mbit down and 2.8Mbit up. Only IPv4 is enabled on this provider in my current configuration.
2) Starlink, providing an average of 200Mbit down and 20Mbit up. Stable, but with a lot of jitter. IPv4 and IPv6 are both enabled on this provider in my current configuration.
My DNS settings network-wide are using my local Synology NAS for IPv4 DNS, which uses CloudFlare upstream. My IPv6 DNS is direct to CloudFlare.
My router is currently a Juniper SRX 340. This SRX is configured to use both ISPs in a high-availiability manner, ensuring that I don't lose internet access even if one ISP drops out.
My wireless access points are Juniper Mist AP41 for the indoor pair, and a TP-Link Omada AX1800 for the outdoor unit.
All access points connect to a Trendnet TPE-TG80G 8-port passive PoE switch, which is then itself plugged into the SRX router. 1Gbit wired connectivity throughout.
The network itself is a simple, flat network with no clans. MTU is default. IPv4 is DHCP, IPv6 is stateless autoconfig.
As part of my troubleshooting I have connected cameras directly to the Telus CPE modem/router as well as the Starlink wireless. I recieved the same "the network is unstable" using the ISP-provided factory defaults and using one ISP at a time as I do when the cameras are connected to my own high availability network.
As part of my troubleshooting I have deliberately waited until there was a prolonged period of "network is unstable messages", and then observed each camera individually, and then all cameras simultaneously first through RSTP then through Onvif using third-party applications. The cameras performed flawlessly when viewed in this fashion.
I have also used my Synology NAS to record the cameras using Onvif event detection, and the Synology worked fine during the "network is unstable events".
Given that RSTP and Onvif seem to work when the TAPO app does not due to "network is unstable" messages, my preliminary conclusion is that my local network is fine, the cameras' connection to the local network is stable, and that the cameras themselves are not experiencing some sort of resource constraint that prevents them from operating during the "network is unstable" periods. I also believe it likely that the TAPO app does not connect to the cameras using the local network when access either the livestream or the recordings stored on the MicroSD cards. This would mean that the TAPO app uses a bent pie through TAPO's cloud to connect my smartphone to the cameras, even when both devices are on the same local network.
Taking all of the above I to consideration my suspicions now point to the TAPO cloud as the potential culprit for the "network is unstable" messages. I have tested and troubleshot these cameras and the network meticulously for months. I have connected them in every configuration and networking scenario that my equipment will allow. I even bought a fifth C320WS so that I could set it up in the lab in a controlled environment using a dedicated LTE connection to the internet to rule out my entire network as a cause.
I have had this issue when using the TAPO app on a CAT S62 Pro (Android 11), on a Lenovo Tab P11 Pro Gen 2 (Android 13), and a Ulefone Power Armor 18T Ultra (Android 13).
I am at this point out of ideas to troubleshoot on my end. The problem appears to be some sort of issue on TAPO's end, especially given that the app appears to use a bent pipe to talk to the cameras. Given the intermittent and unpredictable nature, I suspect a resource constraint issue on TAPO's side, but that is nothing more than informed conjecture.
I would appreciate any ideas you have for additional troubleshooting steps you would like me to take before we can get a ticket created to investigate possible issues that may exist on TAPO's end.
Thank you for your time.