Page 1 of 1
Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 6:56 pm
by CloaknDagr
Cumulus has been working fine for me except for one little annoyance. Every night at midnight it duplicates the date from the day before on the last line. I get an LED Error flashing and when I open it it says to edit dayfile.txt. It's always the last line and it always has the date from the day before which needs to be manually edited. After I edit by replacing (usually just one number except at midnight the last day of any month) the wrong entry Cumulus works good for another day.
There are random, sporadic issues like "sensor data lost" and such but I'm pretty sure they're just normal communications problems and they don't happen very often.
But this dayfile.txt problem occurs every day, well night actually at midnight. Most of the time I'm up and I've seen it throw that error seconds after exact midnight.
So if anyone could give me a hand finding this and correcting it I'd be grateful.
I always run the latest stable release also. I figured I'd wait for a release cycle to see if it was a bug that got fixed but the error carried over from 1.9.3 to 1.9.4
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 7:32 pm
by steve
Please zip up the diags folder and attach it.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 7:45 pm
by CloaknDagr
Here you go. Thanks for having a look.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 7:58 pm
by steve
Thanks - that's interesting, it does look like a bug. I suspect that you're seeing it because of your 1-minute logging interval. I'll see if I can work out why it happens.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 8:07 pm
by steve
Ah, hold on. You are editing dayfile.txt every day so that the latest entry has today's date? The latest entry in dayfile.txt should always have yesterday's date. When you change the date, you are changing the entry for yesterday to look like today's entry. Then at midnight Cumulus creates the genuine entry for that day, and you end up with two entries for the same day, which it spots when it does the NOAA report.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 8:23 pm
by CloaknDagr
Every day it throws an error "duplicate entry at line ..." and gives the last line of the dayfile.txt. The only way to get it to clear the error is to change that date to one higher. I can't change it to one lower, it's still a duplicate then.
So, unless I delete a line of data this is the only way to clear the error...
Should I delete a line of data?
I'm a Certified Advanced NWS SkyWARN Spotter so it's kind of important to me to get this working right. If you could tell me how to correct this problem I'd appreciate it.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 8:29 pm
by steve
Yes, you will have to delete an entry. Presumably at some point you had some kind of issue which resulted in two entries being logged in dayfile.txt for the same date. It's impossible to say now why that occurred, but presumably you saw the error message and changed the date on the last line to the current day, rather than removing the extra entry. Since then, all of your entries in dayfile.txt presumably have a date one day later than the day to which they apply.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 16 Mar 2014 9:35 pm
by CloaknDagr
OK, here's what I did-
1. Closed Cumulus via File/Exit.
2. Opened the Dayfile.txt file from "C:\Cumulus\Data\dayfile.txt" location.
3. Deleted the last entry (the entire last, bottom line) of data in that file. That line had today's date at the beginning.
4. Saved dayfile.txt to it's default location.
5. Launched Cumulus.
Cumulus came up just fine after that so we'll see if the error comes back at midnight. At midnight it should post a new line of data with today's date (which will be yesterday at that time) so the error should not repeat.
I didn't know that the dayfile.txt file was data only up to the previous day, I thought it included data for today up to the last data read. There's an awful lot to sift through for documentation and the forum posts, and I did read all the FAQ's though I may have missed or forgotten something. I'm sorry if this was covered elsewhere and if so I goofed.
I'm an IT pro so tinkering with computers is nothing exciting but I'm new to owning a weather station and I am finding it to be very interesting and enjoyable.
Thank you for taking the time to help me with this.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 25 May 2014 5:33 pm
by aznetcowboy
CloaknDagr wrote:Every day it throws an error "duplicate entry at line ..." and gives the last line of the dayfile.txt. The only way to get it to clear the error is to change that date to one higher. I can't change it to one lower, it's still a duplicate then.
So, unless I delete a line of data this is the only way to clear the error...
Should I delete a line of data?
I'm a Certified Advanced NWS SkyWARN Spotter so it's kind of important to me to get this working right. If you could tell me how to correct this problem I'd appreciate it.
I am running into the same issue as CloakDagr. And I am also a Certified NWS SkyWARN Spotter, so accuracy is important. I have been fighting this for the past several days. Today it is line 510. I have used the Cumulus Edit, the Cumulus Toolbox Edit, Notepad++, and MS Notepad to remove line 510. Oh yes, for the last three edits, Cumulus has been exited. I am using version 1.9.4 build 1087 of Cumulus.
In every case, line 510 reappears looking exactly like it did before. In each case it gives me a low temperature of zero (not going to happen here in Tucson!) and a zero low pressure and wind. A reading of zero for the humidity is possible (The NWS here recoprded that as the official low humidity last Monday, believe it or not), but not a low temperature of zero.
How do I get around this error?

I have included the dayfile.txt as an upload.
Re: Dayfile.txt problem consistently repeats.
Posted: Sun 25 May 2014 5:37 pm
by steve
Please zip up the diags folder and attach it.
Note that last line of your dayfile.txt looks like this:
24/05/14,26.0,258,13:10,56.3,05:24,83.5,15:12,29.659,06:42,29.859,09:44,0.00,09:44,0.00,78.5,38.2,7.0,13:20,12,05:51,26,09:44,0.000,0.0,80.8,15:12,78.2,13:46,56.3,05:04,0.00,09:44,0.0,09:44,37.4,09:56,25.7,18:17,315,0.0,0.0,0,09:44,0.0,09:4424/05/14,9.0,249,09:52,0.0,09:51,79.7,09:51,0.000,09:51,29.823,09:52,0.00,09:51,0.00,63.7,0.3,4.9,09:52,0,09:51,24,09:51,0.000,0.0,79.7,09:51,76.5,09:51,0.0,09:51,0.00,09:51,0.0,09:51,39.7,09:51,31.9,09:51,276,0.0,0.0,0,09:51,0.0,09:51
So there are two entries (with the same date) on the same line. Presumably when you edited the file, you removed the end of line markers for the last line, so the next entry which was added got appended to the same line.
Note also that the OPs problem was caused by editing the date for the last entry to be today's date, when it should be yesterday's date. Presumably you haven't been doing that, so you don't have the same problem.
Re: Dayfile.txt problem consistently repeats.
Posted: Tue 27 May 2014 4:19 pm
by aznetcowboy
I have done as suggested. Funny thing, when I first posted my message, I was going crazy over the problem. After posting it, everything started working again. I was going to send you the diags folder data, but there wasn't was anything in the folder. Turns out the diagnostics option was turned off. I don't remember disabling that option, but it was, so I re-enabled it. Dumb mistake on my part.
Well, today it started in again. So I now had something to send in to you. It is attached.
Re: Dayfile.txt problem consistently repeats.
Posted: Tue 27 May 2014 4:54 pm
by steve
Thanks - you can't turn the diags logging off, by the way, it's always on. You may be confusing it with the debug log, which should normally be left switched off.
When you started Cumulus up this morning and it tried to download the overnight data, it was told (either by the console or the Davis DLL) that there wasn't any to download:
5/27/2014 08:08:30.715 : VP2: Downloading history from 5/26/2014 9:40:00 PM
5/27/2014 08:08:30.715 : y=2014 m=5 d=26 h=21 m=40
5/27/2014 08:08:44.016 : VP2: DownloadData_V = -32701
5/27/2014 08:08:44.016 : VP2: GetNumberOfArchiveRecords_V = -1
5/27/2014 08:08:44.016 : VP2: No history records to process
In yesterday's diags file, there are a lot of entries where the DLL was unable to reconnect to the console after its periodic disconnect. Presumably you have the WeatherlinkIP uploading to weatherlink.com? Perhaps the upload is taking a long time. This could explain why the DLL was unable to download any logger entries.
I'm afraid there's not much I can do about this, the Davis DLL handles the communication with the station, and the poor design of the WeatherlinkIP makes things very difficult. What you could do is watch the Cumulus screen when it starts up to see if it has downloaded any logger entries (it display the number downloaded in the status bar, at the bottom left). If it didn't download any archive entries, close it and look for the backup folder which it has just created (you can tell which it is from the name of the folder or the Windows timestamp). Copy the files from that folder into the Cumulus data folder, and start Cumulus up again; this will cause it to have another try at downloading the logger entries.
As a side note, you might want to consider reducing the logging interval in the WeatherlinkIP from the 30 minutes that you have it set to, or even leaving your PC on 24/7. You would get much better accuracy that way.
Actually, regarding that last point - you have the station logger set to 30 minutes and Cumulus set to 10 minutes, despite the advice to have them both the same. Normally this causes the logger to spew out its entire contents (almost) every time, but perhaps with your logger for some reason it causes no entries to be downloaded. It's interesting that yesterday entries were downloaded successfully, and the last Cumulus logger entry was on the half hour, so coincided with a station logger entry.
Re: Dayfile.txt problem consistently repeats.
Posted: Thu 29 May 2014 4:46 pm
by aznetcowboy
steve wrote:Thanks - you can't turn the diags logging off, by the way, it's always on. You may be confusing it with the debug log, which should normally be left switched off.
{SNIP}
Actually, regarding that last point - you have the station logger set to 30 minutes and Cumulus set to 10 minutes, despite the advice to have them both the same. Normally this causes the logger to spew out its entire contents (almost) every time, but perhaps with your logger for some reason it causes no entries to be downloaded. It's interesting that yesterday entries were downloaded successfully, and the last Cumulus logger entry was on the half hour, so coincided with a station logger entry.
You are correct in that I had the debug confused with the diags file.
As to keeping the system online 24/7, that's an argument you'd have to win with my Wife. I let her pay for the utilities and I pay the mortgage. She's on my case more than I hate to count about the electric bill. Now that it is summer and temps are averaging almost 100 degrees Fahrenheit, with the A.C. running as much as it does, I will be hearing those complaints again.
I have modified the datalogger to download more frequently, so let's see what happens <fingers crossed>. If the problem persists, I will contact Davis and deal with their tech support. I just hope they are as good to deal with as you are.

Re: Dayfile.txt problem consistently repeats.
Posted: Thu 29 May 2014 5:10 pm
by steve
aznetcowboy wrote:I have modified the datalogger to download more frequently, so let's see what happens <fingers crossed>.
You've set it to the same as Cumulus, hopefully?
Re: Dayfile.txt problem consistently repeats.
Posted: Thu 29 May 2014 5:24 pm
by aznetcowboy
steve wrote:aznetcowboy wrote:I have modified the datalogger to download more frequently, so let's see what happens <fingers crossed>.
You've set it to the same as Cumulus, hopefully?
Yes I did. I guessed if 30 to 10 minutes makes a difference, then anything except 10 to 10 minutes wouldn't help.