Cumulus Support

Yes, that looks OK. The only other explanation is that the station was returning an invalid value for the temperature reading.

I will check it further.
I have not seen it 1 week ago, when I worked at my start/stop script and restarted Cumulus to test it. Yesterday, Cumulus ran for 4 days without a restart and these problems began. Now I will check tomorrow.
I modified my Start|Stop script to accept Parameter -r (restart) to use is also in cron to restart it once per day, if this solve my problem.

As mentioned yesterday,
I Restarted Cumulus yesterday and had no further "flat values" until now. I will further Check it

Hello Steve.
since I am restarting CumulusMX frequently, I did not see the "flat" temperature graph anymore.
Not sure, if the restart fixed something, but beginning with next week, I will again stop restarting cumulusMX each day, to find out, how many days I need top reproduce.

I have another question, On my CumulusMX graphs I sse some strange lines. It was around the time, when I worked on my script yesterday where I restarted cumulus many times for testing.
I am talking about the back and forward in the lines.

Nice patterns! It looks like you have multiple entries for the same time period, or a problem with the timestamps. Perhaps your station clock is wrong? Or it could be a bug in MX. If you zip up the mxdiags folder and attach it, together with your May log file, I'll take a look.

Is it possible to run two copies of MX Steve, as I suppose that would cause this.

Yes, that's quite a likely candidate too. And I can't remember whether the setting to stop multiple copies is working in 3024. It's in 3025 - coming quite soon, hopefully.

water01 wrote:Is it possible to run two copies of MX Steve, as I suppose that would cause this.

I am sure, I am not running 2 copies of Cumulus MX
because I am starting my cumulus by a script creating PID file and I am stopping cumulus by

MONOPID=$(pgrep mono && echo) will get the PID of MONO running
sudo screen -X -S "CumulusMX" stuff "^C" Will send a CTRL+C to the Cumulus Screen Name
sleep 5 Giving 5 Seconds to Sumulus to stop
sudo kill -5 $MONOPID before sending a KILL to MONO
rm /run/$SERVICE.pid removing the PID File

so I am 99% sure, that there are not mulitple instances
Btw: I am running CumulusMX on Debian not windows

steve wrote:Nice patterns! It looks like you have multiple entries for the same time period, or a problem with the timestamps. Perhaps your station clock is wrong? Or it could be a bug in MX. If you zip up the mxdiags folder and attach it, together with your May log file, I'll take a look.

Edit: I think I found some interestiong thing:

tail -f Mai15log.txt
than I restarted cumulus with this reult in the MAY Log File

20.05.15;15:35;15,4;58;7,2;5;8;123;0,0;1,0;1013,7;1171,2;20,5;41;3;15,4;15,4;0,0;0;0,00;0,00;13,7;995;0,0;180;0,0;1,0
20.05.15;15:40;15,8;57;7,3;3;7;160;0,0;1,0;1013,7;1171,2;20,6;41;0;15,8;15,8;0,0;0;0,00;0,00;14,6;985;0,0;0;0,0;1,0
20.05.15;15:45;15,8;55;6,8;3;8;74;0,0;1,0;1013,7;1171,2;20,6;41;8;15,8;15,8;0,0;0;0,00;0,00;14,6;975;0,0;67;0,0;1,0
20.05.15;15:57;15,2;61;7,7;5;5;45;0,0;1,0;1013,8;1171,2;20,4;42;5;15,2;15,2;0,0;0;0,00;0,00;13,7;949;0,0;45;0,0;1,1
20.05.15;15:50;15,7;56;6,9;5;12;87;0,0;1,0;1013,7;1171,2;20,6;41;6;15,7;15,7;0,0;0;0,00;0,00;14,0;964;0,0;157;0,0;1,1
20.05.15;15:55;15,7;55;6,7;4;12;98;0,0;1,0;1013,8;1171,2;20,6;41;0;15,7;15,7;0,0;0;0,00;0,00;14,1;954;0,0;0;0,0;1,1
20.05.15;16:00;15,4;55;6,4;2;6;110;0,0;1,0;1013,8;1171,2;20,6;41;0;15,4;15,4;0,0;0;0,00;0,00;14,3;942;0,0;0;0,0;1,1
20.05.15;16:57;15,6;55;6,6;0;0;0;0,0;1,0;1013,8;1171,2;20,6;41;0;15,6;15,6;0,0;0;0,00;0,00;14,8;798;0,0;0;0,0;1,1
20.05.15;16:10;15,2;55;6,2;0;0;0;0,0;1,0;1013,7;1171,2;20,5;41;0;15,2;15,2;0,0;0;0,00;0,00;14,3;919;0,0;0;0,0;1,1
20.05.15;16:15;15,1;55;6,1;0;0;0;0,0;1,0;1013,7;1171,2;20,6;41;0;15,1;15,1;0,0;0;0,00;0,00;14,2;908;0,0;0;0,0;1,1
20.05.15;16:20;14,9;56;6,2;0;0;0;0,0;1,0;1013,8;1171,2;20,6;41;0;14,9;14,9;0,0;0;0,00;0,00;14,0;895;0,0;0;0,0;1,1
20.05.15;16:57;15,6;55;6,6;0;0;0;0,0;1,0;1013,8;1171,2;20,6;41;0;15,6;15,6;0,0;0;0,00;0,00;14,8;798;0,0;0;0,0;1,1
20.05.15;16:25;14,9;58;6,7;0;0;0;0,0;1,0;1013,9;1171,2;20,6;41;0;14,9;14,9;0,0;0;0,00;0,00;14,1;883;0,0;0;0,0;1,1
20.05.15;16:57;15,6;55;6,6;0;0;0;0,0;1,0;1013,8;1171,2;20,6;41;0;15,6;15,6;0,0;0;0,00;0,00;14,8;798;0,0;0;0,0;1,1

the strange thing is, I restarted it 4 times, and every time when I restarted it, the time jumps to xx:57:15

Meanwhile, Ignore LaCrosee clock is checked, System Date is: Mi 20. Mai 16:31:28 CEST 2015 , Station Time is set to the correct time.
In the MXDiags Log, I can see this entry:
2015-05-20 16:28:12.835 Last update time from today.ini: 20.05.2015 16:57:00
2015-05-20 16:28:12.836 Read today file: Date = 20.05.2015, LastUpdateTime = 20.05.2015 16:57:00, Month = 5

Why is cumulus stating, that the last entry in the today.ini is in the future ?

and indeed, I restarted it again at 20:02 and looking into the today.ini

After I stopped Cumulus:
[General]
Date=20.05.2015
Timestamp=20.05.2015 19:55:00
CurrentYear=2015
CurrentMonth=5
CurrentDay=20

After I started it again 1 Minute later:
[General]
Date=20.05.2015
Timestamp=20.05.2015 20:57:00
CurrentYear=2015
CurrentMonth=5
CurrentDay=20

Finally, when Cumulus is starting, it is always writing the next full hour -> xx:57 into the today.ini