Local Dashboard updates stop on Nano SD

[Snapper]

So to be clear, when it stops, the MB Nano webUI is working normally, but shows no internet?

If so, this is an issue with the router or repeater, not MB.
Getting the webUI on MB is showing the WiFi and local connectivity part is working, but MB thinks internet isn’t available because it can’t reach outside the local LAN.

Do you have access to the router?
I would be interested in what the DCHP lease expiry is set to, and if it matches the times when connectivity is lost.

[thunderstorm]

I haven't checked whether it's saying no internet and the exports aren't complaining that there's connection failure. They're just not scheduled anymore. Also the buffer is filling up, which shouldn't happen just because MB thinks there's no internet.

Besides that the repeater is working properly and all other devices connected to it are working fine and have internet access.

The dchp lease time is 10 days, but the hangups are irregular and I can't see a correlation.

[Snapper]

Ok, fair enough.

You can see when connectivity is a problem because you get the message about Access point temporarily enabled in the log.

I wouldn’t discount the repeater just yet though - i work with Cisco network kit and when I brought some of their AP’s home from my test lab, i could see the MB Nano bounce and disassociate frequently. Caused the Ciscos to reboot at least daily, but when MB-Nano is connected to my home Ubiquiti AP’s, I don’t see an issue.

Have you got band-steering enabled?
Might be worth turning it off to see if that helps.

[thunderstorm]

I'll have a look.

Interestingly there's only
system (24.04.2020 07:20:48): Access point operation disabled 322665 secs after boot.

in the logs. No other entries +/- 25min

[thunderstorm]

It just happened again
- dhcp lease is still valid for >3 days
- band steering and roaming is disabled

The really bad thing is that the buffer is running full and I might lose data if I don't detect it in time (esp. during night). So I'm still wondering why that part (e.g. db-writes and service-scheduling) stops and works fine after the AP rebooted.

[pwilman]

Hello,

I get this issue from time to time as well. My buffer fills up to 100% and then all services stop uploading. Its probably no co-incidence, but I also get quite a few records in the system log on disconnecting from WLAN, and the associated access point disabled message.

Boris has looked into this a few times for me. The latest suggestion from him was to reduce the upload to the weather34 template from every 10 seconds to every 30 seconds. That way it is uploading less frequently, so less likely to be uploading when I have connection issues (which is apparently when the buffer fills up).

Having done this, I went for 69 days without the buffer filing up.

Not sure if this is the same issue my nano SD connects directly to my router? Please ignore if not related.

Paul

[rks]

I use 1 minute updates for both Weather Underground and Local Dashboard over a direct link to my router and all is working well with no problems at all. Very occasionally see 2 items buffered but usually just one, which I assume is normal. I still would like to use a wifi repeater to allow me to reposition the AWS but as they say ,if it ain’t broke don’t fix it!