Hi
Three times today the Base Unit has gone down, or is not contactable, ai am not sure which. Nothing has been or was done prior to this. It has been working fine. Checking the CMS logs the same error is reported on all three occasions …
I do not know what …mqtt.php does. The card is standard as issued when purchased. Another observation is that the Tx would seem to be OK as the Admin entry says some 300 odd lines were waiting to be written.
How can this be stopped ? Is there an auto-reboot facility if that’s what it needs. After cycling the power, normal service is resumed.
[edited by MOD] Please format code/logs etc with Ctl-E
That happens all the time, not an issue. It was disconnected from the MQTT Broker for less than a tenth of a second.
By Base Unit do you mean an emonBase (i.e. an RPi)?
Is it connected by Wi-Fi? If so I’d start there particularly if using an ISPs router and runing in Dual Band mode (shared SSID for both frequencies). Have you reserved the IP Address?
So I think you are confusing different things. The emonBase, by default, stores the data in memory, then writes it to disk as a batch - that is those 300 readings.
Yes. Well I’ve had the equipment about a month now. All has been well until the last few days when the problem as described started. In the last couple of days it has continued. However I have refrained from rebooting and waited. Literally refreshed after a timeout in Safari. Refreshed again and then magically the log in screen comes back and it works fine …… for a while. Even more oddly, while waiting for Safari on my iPad to timeout, I went to my Mac where I found it was working fine !!! Then going back to my iPad, it also had started working fine !!!
It feels like sessions/web server or networking. Need to find logs to analyse more.
No so I’d notice.
I haven’t tried SSH but Fing on my Mac would ping it fine when it was not responding on Safari.
Ordinarily just emonpi. If that fails then the assigned IP address. Failure to contact has occurred using both so that rules out DNS. Are the web server logs usual Apache on a Pi ?
Will do. This last couple of days I notice that if I am patient and wait for the timeout in the browser, then refresh and wait for the timeout again, then refresh again the log in screen will appear … normal service is resumed. I’m starting to notice this is repeatable, so am counting and will investigate further.
Previously, I was always using ‘emonpi’ in the Safari browser. Most times it would connect ok. Some times it would pass the term emonpi to google as a search term, so likely a DNS problem in this instance. Occasionally, there would be no response at all. For several days now I have been using ‘emonpi.local’ without any failures. If I don’t post again, then this was the solution. Thx @borpin