B3032 - NOAA reports discrepency

[mcrossley]

Steve

I notice there are some differences in the November report and the 2015 report.

The monthly report has the mean temp for the month so far as 9.0, the 2015 report 9.1
The monthly report has the total heating degree days as 161.7, the 2015 report 154
Similar for total rainfall....

If you knock off the last days heating deg day (7.1) from the monthly figure, you get the yearly report figure, so it looks like the yearly report may not be including the last available day of the month in its calculations? The same applies to the rainfall m=104.2 y=101.6 and... m-lastday[2.6] = y

You don't say you changed anything in b2034 for these reports so I assume it is the same. I have just updated to b3034 and I'll check it again tomorrow.

[steve]

Mine look OK. I use an 0900 start of day, but I don't think that should make any difference. Perhaps your annual report didn't get updated today for some reason? Any errors in the diags log?

[mcrossley]

Hmm, on the web server both files show update times 26/11/2015 01:05:00

The diags file is a bit odd though, it has the upload at 00:05 which matches the file time. It starts creating the annual report at 00:01:12, but doesn't save it until 00:07:05 - after the upload! ...

Code: Select all

2015-11-26 00:00:00.252 Hour changed:0
2015-11-26 00:00:00.255 Calculating sunrise and sunset times
2015-11-26 00:00:00.264 Sunrise: 07:52:46
2015-11-26 00:00:00.265 Sunset : 15:59:19
2015-11-26 00:00:00.292 Tomorrow sunrise: 07:54:23
2015-11-26 00:00:00.301 Tomorrow sunset : 15:58:21
Midnight rain reset, count = 714.2 time = 00:00
2015-11-26 00:00:00.378 === Day reset, today = 26
2015-11-26 00:00:00.380 === Day reset for 25/11/2015 00:00:00
2015-11-26 00:00:00.381 NOAA reports will be uploaded at next web update
2015-11-26 00:00:00.389 Dayfile.txt entry:
2015-11-26 00:00:00.391 25/11/15,19,336,23:32,5.3,00:54,9.4,13:40,1008.20,00:31,1016.86,23:58,6.2,22:33,2.6,7.9,123.3,9,13:15,88,13:03,95,00:00,0.30,0.0,9.4,13:40,7.8,13:47,2.4,00:24,1.2,00:22,3.2,00:24,8.5,13:46,4.5,00:54,316,7.6,0.0,127,11:59,0.8,11:51
2015-11-26 00:00:00.392 Dayfile.txt opened for writing
2015-11-26 00:00:00.394 Writing entry to dayfile.txt
2015-11-26 00:00:00.641 INSERT IGNORE INTO daydata ....
2015-11-26 00:00:08.265 MySQL: 1 rows were affected.
2015-11-26 00:00:08.266 Raincounter = 714.2 Raindaystart = 711.6
2015-11-26 00:00:08.273 Rainyesterday (calibrated) set to 2.60000000000002
2015-11-26 00:00:08.275 RainDayThreshold = -1
2015-11-26 00:00:08.278 rdt1000=200 ry1000=2600
2015-11-26 00:00:08.279 Yesterday was a rain day
2015-11-26 00:00:08.279 Consecutive rain days = 3
Raindaystart set to 714.2
2015-11-26 00:00:08.280 StartofdayET set to 582.168
2015-11-26 00:00:08.324 Writing today.ini, LastUpdateTime = 25/11/2015 23:55:00 raindaystart = 714.2 rain counter = 714.2
2015-11-26 00:00:08.442 Creating NOAA monthly report for 25 November 2015
2015-11-26 00:01:12.791 Saving monthly report as Reports/NOAAMO1115.txt
2015-11-26 00:01:12.813 Creating NOAA yearly report
New monthly record, month = 11: 2015-11-26 00:02 104.400 "Highest monthly rainfall" 2015-11-25 00:00 104.200
2015-11-26 00:05:00.072 Writing log entry for 26/11/2015 00:05:00
2015-11-26 00:05:00.117 Written log entry for 26/11/2015 00:05:00
2015-11-26 00:05:00.271 Writing today.ini, LastUpdateTime = 26/11/2015 00:05:00 raindaystart = 714.2 rain counter = 714.4
2015-11-26 00:05:00.307 INSERT IGNORE INTO fulldata ...
2015-11-26 00:05:00.342 MySQL: 1 rows were affected.
2015-11-26 00:05:01.555 WU Response: OK: success

2015-11-26 00:05:32.287 Uploading NOAA reports
2015-11-26 00:07:05.712 Saving yearly report as Reports/NOAAYR2015.txt
2015-11-26 00:07:05.728 Executing daily program: EndOfDayTasks.sh params: &
2015-11-26 00:07:05.903 === Day reset complete
2015-11-26 00:07:05.919 Now recording data for day=26 month=11 year=2015

I suspect it may be delayed by my end of day script running?

[mcrossley]

No, scrap that, the end of day script is run after everything else.

[steve]

For some reason your annual report is taking a very long time to produce - over 5 minutes. The next ftp update comes around before it's complete, and because it's already set the flag that says the reports need uploading, it uploads them. I'll move the setting of the flag, so the update won't be done until the reports are complete. I don't know why the report is taking so long, mine takes about 20 seconds on my Pi2.

[mcrossley]

Looking back most days it takes 1m 10s on my Pi Model B, but there has been the odd day where it has taken 6 to 7 minutes. The only job I run is kicked off by Cumulus as a EOD task, and I can't see any cron jobs that run during that time, one was scheduled at 00:05, so I have moved that to a bit later.

Anyway moving the flag to after the file creations should fix it.