Logging to local emoncms on RPi, has worked pretty well for a long time. I recently found that some feeds on emoncms stopped updating Feb 29, while others continue no problem. The ones that don’t update show as Inactive in the emoncms list, the others have a time in seconds or hours. All the feeds show reasonable values in the Feeds list so seems like data is coming in for them. Could suspect time stamp problems but why only on some feeds?
I ran updates on emonpi and rebooted, no change. I restarted iotawatt, could not connect, suspect a separate problem, had to reboot router to get it back, but also no change to the feed logging problem. I do see an error message in the log: EmonService: get input list failing, code: -4 after the restart, suspect could be related. Here’s the message log, any clues here?
Note: router was rebooted about 20:55. Before that, you can see in log that iotawatt was using a private IP address, so did not get DHCP IP. Router showed iotawatt as active, and using the same good IP address it always has, but could not ping or connect, of course, because iotawatt thought it had a different IP.
iotamsgs.txt (9.7 KB)
It could be WiFi issues or the emoncms server. My money is on the emoncms server. The other possibility is that there is something in your environment that interferes with the WiFi from time to time. I see that you are using channel 10. People seem to get good results with 1,6 and 11 here in the US, which is what your timezone suggests.
Regarding the feeds nor updating, it is definitely not the timestamp. The way the protocol works, all of the measurements for a single interval are sent with a single timestamp. In addition to examining the emoncms last update time for the feeds, take a look at the same data for the inputs. If they all show as current, then you might take a look at the process list for the input that should be posting to the feed(s).
Hmm, good point about the emoncms inputs. I glanced at them but did not pay enough attention as I thought they “matched” the feed issue. But now I see the problem - there are duplicate inputs for the ones not updating. There is data coming in for those inputs, but there is no process list, and they are identified (ARG) only by number, like Iotawatt:2, where the matching duplicate is Iotawatt:MainA. The input definition for both the original and the duplicate show Iotawatt:2, where the original has the MainA in teh description field, the duplicate description field is blank. Not yet clear to me how to fix it, ideally without losing too much history. Which one to delete? Maybe I delete the original and just enter the feed process to the live duplicate. Not sure what happens if I delete the live duplicate. Onward! Thanks for the reply…