Correct way to halt data logging on Vantage Vue?

[philpugh]

I recently had a 'set of faulty readings of Temp and Humidity on my Vantage Vue - this is the second time in it's life (bought 2009).

So it was time for a full strip down, clean and re-assembly.

What I had done at the time of discovering the problem (during a wet evening) was to close down CumulusMX (and halt and de-power the Pi running it). I unplugged the Davis Console and removed the batteries.

The next afternoon, after hopefully fixing the problem, I started up the Davis console - checked I was getting sensible readings - they were in close agreement with my other station (Froggit WH3000SE) so I connected up the the Pi and restarted it and CumulusMX. What I was expecting was a gap in the data - both in the JSON and also the MySQL but instead I had a set of continuous data!

Not a major problem to fix the erroneous data but my question is where did it come from? All recorded data (JSON and MySQL) for the 'off' period was the same - no variation in any item.

I have attached the startup Diags for this and it shows a realtime update of over 2000 rows after the initial startup. Any ideas?

[mcrossley]

No idea where that realtime insert of 2192 rows came from.
Are you using the built-in table update or a custom command?
Any customisation of the table in MySQL?

The realtime SQL command is just doing a single row insert.

[philpugh]

Mark,

I am using the standard MySQL.

Today I had to redo everything as there were glitches in the temp/humidity again overnight. I think the unit is on its way out sadly.

I modified my website to take data from the Froggit system I have for the main display.

I then restarted the Davis system and pi just as before and everything was as I would have expected - CumulusMX read the old data in the logger but didn't process any of it.

Don't know what caused the glitch the first time!

Cheers

[mcrossley]

The two thousand odd rows affected will be it deleting the old entries. I'll change the messages to "rows added" and "rows deleted" rather than both being "rows affected".