I finally decided to postpone the purchase of a new wifi access point. In the meantime, I remembered I had a spare router on a drawer so I plugged it in and now I have two separate wifi access points, one for my IoT devices and the other for all the rest. No more conflicts.
To anyone experiencing network problems, I recommend reading Shelly Networking Best Practices
Thanks to everyone who wrote with suggestions.
After reading the Shelly Networking Best Practices I think I have the problem identified. The sad truth is my old access point just cannot handle this many devices, and "old" being the operative word.
Until now, this access point served me well. Most of my critical devices (computers and multimedia devices) are wired, so I never noticed any trouble. That is until now... it seems there's a (not so large) limit in the amount of simultaneous wifi devices the AP can handle.
When I went to look at the specs of my old access point, a Netgear DGND3700 v1, I realize how old it is (802.11N from 2011). A whole decade went by since I installed it, so it seems it is time for an upgrade.
I'll report back when I do.
Hi acrowe, one of them may be prompt to be the first one to fail, but nothing conclusive.
As mentioned, it is not failing right now. However, if the problem reappears, one thing I will try is physically separate them. They were practically touching each other.
At the moment, it is working fine and now it is not failing anymore. But to answer your questions:
No, it was not falling back to AP-mode. In fact, the device kept working, it just would not answer. I know this, because if I constantly ping the non-responsive device, as soon as I click on reboot on one of the other devices, the ping starts receiving the packets back. Also the ping I had running on the rebooted device stops and will not resuscitate until I reboot any of the other two devices.
I really wish it starts failing again so I can get more information by sniffing the network traffic with wireshark. I'll post any results if the setup begins failing again.
I'm not using the backup wifi functionality.
I'm sure the pool is not running out of IP's. In fact, to rule out a dhcp problem, I have removed those IP addresses from the dhcp pool and I am setting the addresses statically.
I'll report back if it starts failing again.
Thanks, I ruled out a duplicated IP already, but I will check again. I should note that I have all my IP addresses properly documented. In fact, I keep my dhcp leases in my router (pfSense) tied to the mac address of each device in my network (not only shelly's, TVs, computers, printers, etc). This is to ensure IPs given are consistent across reboots.
The router would not let me assign the same IP to two different mac address, so I'm covered there.
Although the dhcp server can provide IP's dynamically, it does so from a pool of address in a different range. So if I accidentally forget to add a device to the mac-IP list, the IP given to such device will never conflict with the ones I setup up.
Another reason to rule out a duplicated IP is that, even after turning off the unresponsive device, pinging to the IP address returns no result. If there were a different device using the same IP I should get a response. There's the unlikely possibility a device with a duplicate IP simply does not have ping enabled... but that is very unlikely in my setup.
I should also add that after a few hours of operation, all three devices became responsive simultaneously. However, turning them off for a while brings back the problem.
Hi, I'm having a weird problem that I just cannot figure out.
I have 3 shelly RGBW2 devices. All three share the same power supply, which means they all turn on exactly at the same time. I mention this, in case it is relevant to the issue.
The problem is, only two devices are responsive. The third device does not respond to ping, and the webUI never loads. Funny thing is, if I reboot any of the working devices, the unresponsive one immediately starts working on its own. However, now the rebooted device is the one not responding. Again, if I reboot any of the other two, the faulty one fixes itself but the rebooted one now does not respond. This is consistent and I can continue to eternity, always seeing the same result. Only two devices are responsive at the same time.
Things I tried, but had no effect:
* I put my wireless router very close, in case signal wasn't strong enough.
* I changed from DHCP to static IP, in case the dhcp sever could not handle multiple simoultaneous lease requests.
* I verified all have different IP addresses. They do: x.x.x.110, x.x.x.111, x.x.x.112
* I used a different browser for each device (running out of ideas already 
).
Some additional information:
* I have MQTT enabled.
* I have the latest non-beta version in all devices: 20201124-092159/v1.9.0@57ac4ad8
I cannot shake the feeling I'm missing something very obvious.
Anyone experienced something similar?
