Communication with emoncms.org has stopped!


#1

Since installation last October communications between IoTaWatt.local and emoncms.org has faltered once before. On that occasion it was solved by changing the router channel setting from ‘auto’ to a specified channel number. However, in the evening on last Saturday (17th), communications between IoTaWatt.local and emoncms.org stopped and has been non existent ever since. Changing the router channel settings to several different channel numbers as well as trying ‘auto’ has not cured the problem this time.

I wanted to upload a copy of the SD Editor iotamsgs.txt file in the hope you might be able to diagnose the issue, but I have not found a way of presenting it in the required graphic format. Any suggestions on the best way forward will be gratefully received.


#2

C[quote=“OddJob, post:1, topic:45”]
I wanted to upload a copy of the SD Editor iotamsgs.txt file in the hope you might be able to diagnose the issue, but I have not found a way of presenting it in the required graphic format. Any suggestions on the best way forward will be gratefully received.
[/quote]

The msgs file would be helpful, I think there’s a way I can allow uploading it but I have to look into that. I’m away for a couple of days so can’t do that now.

Couple of question:

Is the unit still logging? The led should be blinking green rapidly and you should be able to see the current data in the status display as well as graph it with the local graph app.

If that works then could you post a screenshot of the status display with all of the sections open (especially the data logs section).

Thanks.

A screenshot of the end of the messages file as displayed in the app would be helpful as well.

Update, should be able to upload .txt now


#3

Yes, the unit is fully functioning as you described.
The following is the text I intended to send before finding out that .txt files could not be sent - hence the curtailed first post.
Since installation last October communications between IoTaWatt.local and emoncms.org has faltered once before. On that occasion it was solved by changing the router channel setting from ‘auto’ to a specified channel number. However, in the evening on last Saturday (17th), communications between IoTaWatt.local and emoncms.org stopped and has been non existent ever since. Changing the router channel settings to several different channel numbers as well as trying ‘auto’ has not cured the problem this time.
To stop the continual IoTaWatt communication initialisation process, I changed the Web Service Type to ‘None’ on the Setup Web Server window. That said, I occasionally reset the Web Service Type to ‘emoncms’ along with providing the api key and emoncms userid information to check if the problem has been resolved. Unfortunately this is not the case, that is why I wanted to provide a copy of the SD Editor iotamsgs.txt file in the hope you might be able to diagnose the issue.

Sorry, no go - still graphic files only.


#4

Have you double checked the api key? If ok, can you try not specifying the account so as to use insecure post, just to see if that works.

Where are you located? This looks like a latency problem. Since you can run the configuration app, download the log and graph, your WiFi connection seems adequate. So it must be timing out going to EmonCMS. I’ve seen a bit of that from Australia where the message literally goes halfway around the world.

Hopefully I will be putting a new version in ALPHA next week that has more robust asynchronous HTTP. I am optimistic it will solve the latency problem. I just want to determine if that is your likely problem. It’s troubling that it used to work.


#5

Yes, I made a text document record of the api key and emoncms userid information to avoid possible finger trouble later - so I cut and paste when needed.

Done, success! all the missing records appeared on emoncms quickly. A screenshot of the end of the messages file since initiating emoncms with api key only

I live in the UK about 50 mile north of Southampton on the south coast.
When carrying out speed tests to various locations the usual latency figure is in the order of the mid 30ms.

In case you find it useful the next three screenshots relate to the period just before the communication failure on 17_02_18




#6

I looked over the logs, and notwithstanding that you are up and running, I’m not at all confident I know any more about the problem. Could you give me a little more information?

In the log you appear to have stopped Emoncms posting on the 21st at about 17:20. Then it seems to be restarted unsecure today (22nd) at 11:01. Does that sound correct?

Between those times, Wifi was disconnected and reconnected at 23:18. Was that something that you might have done with the router?

Although it appears upload stopped on the 17th, the log indicates that the last upload before you stopped it on the 21st was on the 18th at 8:02:40, so I’d lik eto get a better look at the log. I messed up allowing txt yesterday, but have set it again and tested it so that should work OK.

Id like to see what would happen if you put the account number back in and go back to secure. I think it will work ok, and one way or the other, it will be helpful information.

This may have nexus with another situation I am trying to debug in async HTTP. My MO is to investigate these types of unexplained issues to the fullest extent. Getting to the bottom of this will make IoTaWatt that much more bulletproof.

Thanks for your help.


#7

Yes.

I do not believe it is due to anything I did.

IoTaWatt SD Card Messages 23_02_18.txt (855.9 KB)

I copied the log file about an hour after I inserted the emoncms userid information on the IoTaWatt Setup Web Server window.

You’re welcome.