Unless somebody has a better idea, I’d be looking for something that happens at 3 am to trigger the failure. You’re talking about emoncms.org (on the OEM server) so is your phone company or Openreach doing a line test at that time, or is your router getting a new DHCP lease, so something like that? Because it doesn’t happen frequently, I’s suspect two or more things are happening and you only get a problem when they clash.
Interesting (or coincidence) that it happened at 3 am local time, not 0300 UTC.
And at 3am precisely BST today (24hrs after losing contact) all restarted.
The emonTx & Huzzah are on a fixed IP. My router is a Unifi Dream Machine. This is the 24hr trace from the router, which shows the performance all of a sudden came good at 3am, having been intermittent since 3am the previous night.
I have MyEnergi & GivEnergy systems logging 24/7 to the cloud and neither of them has had any interruption to data flow, so I think that discounts an ISP problem.
The nature of the trace would lead me to think it was a bad connection at the emonTx end but the 3am outage and restore makes me think its a emoncms server issue.
Next time it happens I’ll try pinging the emonTx and see if the internal path is OK.
Possibly, but my thinking is ‘not necessarily’. What if something caused comms to stop at 3 am, and it only retried 24 hours later? It’s a bit far-fetched, but over the years I’ve learned never to discount anything until there’s solid proof that it’s innocent.
Just to put on record that once again I have experienced an exactly 24hour 3am to 3am outage of my feeds.
I was on holiday when it happened so could not do any worthwhile investigation like pinging.
Recording that again, at 3am, this time on 4th Oct, my emonTx once again ceased to transmit data to the cloud.
This time I was home to do a bit of investigating.
Not possible to ping the device.
Turning the 2 power supplies off and on did not solve the problem.
I had noticed constant flashing from the rear of the emonTx, assumed it was the Huzzah, so powered down, removed and reattached it then repowered.
Unit then worked OK.
I’m now wondering whether some action taken on the grid at 3am, like voltage drop, is causing a power supply issue. Mind you that would not make everything come good 24 hrs later.
In the meantime I’m trying a meatier USB power supply. Last one was rated at 6W.
I have a bit of a thing about Ubiquiti gear. I’m not actually sure it is as good as everyone says. I have odd occurrences with mine that are difficult to explain.
I’d take the router out of the equation. Connect the ESP to a different router if you have it. I have 2 of these little beauties (cheaper from China) for my IOT gear and they are rock solid.
Yes, I’m tending to share your opinion about Ubiquiti kit. So far I’ve convinced myself any problems or inconsistencies are as a result of my having missed or got a setting wrong.
My outages all commenced at 3am then restarted at 3am. I’ve just noticed the Ubiquiti UDM is set to auto update and to check at 3am. I imagine this will involve a restart if there’s an update so this could well be the culprit. I’ve disabled auto updates!
I’d have changed the update to 4AM, because if you start getting dropouts at 4AM, you’ll know for definite. Turning it off, may or may have fixed the issue.
