Cumulus Beta 1.9.3 1057 - Graph problems

[Brutus_RST]

Hello,

As this is my first post here, a short introduction.
My name is Bert Bos and I live in Appingedam, Netherlands. I own a Lacrosse WS2350 weather station and as the software that was deliverd by it was not what I expected, I went searching for oher software. And so I found Cumulus. I am following this forum for quite a while and every problem I encounterd, I could solve my self with the answers on this forum.
My website is http://www.weerappingedam.nl and is based on Joomla. I have customised the webpages that Cumulus provides.

Now the following: I have updated Cumulus from 1.9.2 to 1.9.3 beta 1052 while I had the Twitter updates not working anymore.
A new update was released, 1057, and so I updated Cumulus again. Now my Trend graphs are not working properly anymore. When Cumulus restarts, the graph is being reset (quite sure that was ok) and no date before the restart was implemented in the graph. Also I miss some data in the graph from about 0:30 AM till 02:00 AM. As Cumulus wasn't restarted, I went into the december logfile and found all the right data there. But still the graph is missing data.

Another issue is the raincounter. It gives some strange data with the rain from the last 24hr and the total rain for today. Total rain today i.e. 2,5 mm, yesterday 1,6 mm and still the last 24hr rain is 40,4 mm?

If anyone has any ideas, I will be very happy to hear them.

[steve]

Now my Trend graphs are not working properly anymore. When Cumulus restarts, the graph is being reset (quite sure that was ok) and no date before the restart was implemented in the graph.

The most likely explanation is the one in the FAQ: http://wiki.sandaysoft.com/a/FAQ#When_I ... rlier_data

Another issue is the raincounter. It gives some strange data with the rain from the last 24hr and the total rain for today. Total rain today i.e. 2,5 mm, yesterday 1,6 mm and still the last 24hr rain is 40,4 mm?

Quite likely the same problem.

If you can't find a problem in the log file, please zip up the diags folder and attach it, together with your December log file.

Edit: Looking at your temperature min/max/avg chart, it looks like you have duplicate entries in dayfile.txt, which is probably also related to the rain figures. A common reason for this with LaCrosse stations is the clock being wrong, so Cumulus gets data which is incorrectly timestamped (when it's using 'archive' data).

[Brutus_RST]

Hello Steve,

Thanks for the very quick response.

I took a look in the dayfile.txt and there were a few duplicates...
I removed them with Cumulus stopped. Also I attached the Diags and the december log (with the duplicates removed)

The graph isn't updated correctly still, but maybe I need to have more patience.
If you're willing to take a look in the files, it will be much appreciated.

Again, thanks for the quick response.

[steve]

You have some spurious data for 22nd to 30th December in the middle of your data for the 10th. The entries are an hour apart rather than a minute, so they came from the logger; so your station had the wrong date. There's a setting in Cumulus to ignore the LaCrosse clock, which you might want to consider using.

[Brutus_RST]

Thanks Steve,

I think I'm going to use that option.
I'll check the files and get back here if I have an other question, or it is solved.

[Brutus_RST]

Hello Steve,

I completely overlooked the wrong dates.
In the end of the file there were duplicate dates, but also in the 10'th as you mentioned.
There were dates from 22 till the 30th of december in the middle of the data from the 10th.
Removed this, started up Cumulus again (with the option "Ignore Lacrosse clock" and everything is up and running again.

Graphs are allright and also the strange rain reading is gone.

Thanks for the support.