32701 error

[Ben]

I still have the problem that Hans does - sometimes my Davis will respond to my computer, and sometimes not. The biggest nuisance is that if I leave Cumulus running, and the datalogger stops responding, I have to shut Cumulus down. The next time I start Cumulus, Cumulus will only read data from the time I shut it down. An additional problem is that it may take several attempts to restart; Cumulus will give the 32701 error, Weatherlink simply states it has not found the station.
My question is, is it possible for Cumulus to detect that it is not receiving data, and to shut itself down? While it does nothing for the 32701 error, at least when Cumulus is restarted, only a few minutes of data would be lost. Or is it possible to have Cumulus restart itself, and attempt restarting every 10 minutes if it receives the 32701 error? I know this is big revision, but there are a lot of people still struggling with the usb datalogger.
Ben

[steve]

My question is, is it possible for Cumulus to detect that it is not receiving data, and to shut itself down? While it does nothing for the 32701 error, at least when Cumulus is restarted, only a few minutes of data would be lost.

It does this already for IP loggers, as IP connections are liable to fail for various reasons. But it didn't occur to me to do the same for serial/USB connections as I believed they wouldn't present a problem - my serial logger has been operating flawlessly 24 hours a day for four years.

The IP code disconnects and reconnects when it gets a -32701 error; it attempts this once a second for a configurable interval. If it can't reconnect, it attempts to close itself down. I could perhaps look at doing the same (as an option) for serial/USB loggers.

[Ben]

I've noticed that I'm more likely to connect to the datalogger if I remain patient and only attempt it every 10 minutes or so.... trying time after time in succession doesn't achieve any result.
If you decide you will tackle this, I would be happy to be a tester for the revision; my Lacrosse is still online and maintaining full records for me, so I can experiment as much as you like with the Davis.

[steve]

I've noticed that I'm more likely to connect to the datalogger if I remain patient and only attempt it every 10 minutes or so.... trying time after time in succession doesn't achieve any result.

If I do the proposed change, Cumulus will (hopefully) close itself down; you could then use one of the many available 'monitoring' applications to try to restart it periodically. But presumably Cumulus would just put up the 'can't find station' dialogue box on the first attempted restart?

[Ben]

Yes, the change I propose would be for Cumulus to check for the 32701 error, and if it exists, shut itself down. I'm not familiar with the monitoring applications; if they can check if Cumulus is running, and restart it if not, then it would work. A problem may be that Cumulus only displays the error on startup, if the datalogger stops communicating while Cumulus is already running, then it only displays the last time it had a successful download in the bottom right hand corner. Perhaps this may be the check you would run?
I should mention that Cumulus itself does not hang, you only have to press the 'x' button for it to stop.
So for clarity, this is the process I envisage....
Cumulus is run successfully, then stops communicating.
Cumulus checks when last data was read, and if it is more than (say) 2 minutes, shut itself down.
A monitoring program checks if Cumulus is running, and if not, starts it.
Cumulus finds the 32701 error, and again shuts itself down.
The monitoring program then attempts to restart Cumulus on a 10 minute delay, and loop back to the last line.

Does this sound feasible?

[steve]

Perhaps I have misunderstood the problem you are having. You said you were having the same problem as Hans, namely that when trying to fetch data from the Davis DLL, the DLL returns error -32701, so this appears in the diags log:

24.05.2011 18:31:53.359 : VP2: LoadCurrentVantageData_V, error = -32701

The change I am considering making for this situation is as I described earlier - do the same for serial/USB loggers in this situation as it currently does for IP loggers. So if the problem persisted, it would close itself down.

[Ben]

This is the diag for a failed start, everything reads normally until the first mention of 32701. I then close Cumulus manually.


29/05/2011 08:37:08.750 : Rainthisyear from dayfile.txt: 392.100128173828
29/05/2011 08:37:08.886 : Diary log file = C:\Cumulus\data\log.xml
29/05/2011 08:37:09.009 : 8:37:09 AM EWUSB Form Create
29/05/2011 08:37:09.100 : VP2: CloseCommPort_V, res = 0
29/05/2011 08:37:09.100 : VP2: Open connection, type = 0
29/05/2011 08:37:09.103 : VP2: OpenCommPort_V, port = 8 res = 0
29/05/2011 08:37:09.103 : VP2: SetVantageTimeoutVal_V, res = -32701
29/05/2011 08:37:14.018 : VP2: InitStation_V, res = -32701
29/05/2011 08:37:17.046 : MainForm can close
29/05/2011 08:37:17.046 : Showing splash screen
29/05/2011 08:37:17.311 : Exit FormCloseQuery
29/05/2011 08:37:17.311 : MainForm.FormClose
29/05/2011 08:37:17.311 : Terminating threads...
29/05/2011 08:37:17.311 : Threads terminated
29/05/2011 08:37:17.313 : Cumulus closing
29/05/2011 08:37:17.313 : No data read this session, today.ini not written
29/05/2011 08:37:17.340 : Writing cumulus.ini file
29/05/2011 08:37:17.352 : Terminating ftp thread...
29/05/2011 08:37:17.352 : FTP thread terminated


And this is the diag when the problem is encountered after a successful start.

29/05/2011 01:50:00.834 : Writing today.ini, LastUpdateTime = 29/05/2011 1:50:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:51:00.691 : Writing today.ini, LastUpdateTime = 29/05/2011 1:51:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:52:00.533 : Writing today.ini, LastUpdateTime = 29/05/2011 1:52:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:53:00.375 : Writing today.ini, LastUpdateTime = 29/05/2011 1:53:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:54:00.216 : Writing today.ini, LastUpdateTime = 29/05/2011 1:54:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:55:00.058 : Writing today.ini, LastUpdateTime = 29/05/2011 1:55:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:56:00.914 : Writing today.ini, LastUpdateTime = 29/05/2011 1:56:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:56:04.361 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:22.941 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:26.638 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:31.271 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:37.433 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:42.378 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:46.076 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:51.005 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:56:58.400 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:00.755 : Writing today.ini, LastUpdateTime = 29/05/2011 1:57:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:57:04.562 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:08.337 : Error parsing bardata: ''C'' is not a valid integer value
29/05/2011 01:57:13.282 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:20.692 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:29.319 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:35.481 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:41.643 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:57:53.967 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:58:00.597 : Writing today.ini, LastUpdateTime = 29/05/2011 1:58:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:58:01.361 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:58:34.418 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:59:00.439 : Writing today.ini, LastUpdateTime = 29/05/2011 1:59:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 01:59:08.925 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 01:59:12.700 : Error parsing bardata: ''TIO'' is not a valid integer value
29/05/2011 01:59:49.688 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:00:00.296 : Writing today.ini, LastUpdateTime = 29/05/2011 2:00:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 02:00:36.519 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:00:40.295 : Error parsing bardata: ''DQ'' is not a valid integer value
29/05/2011 02:01:00.153 : Writing today.ini, LastUpdateTime = 29/05/2011 2:01:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 02:01:27.141 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:02:01.009 : Writing today.ini, LastUpdateTime = 29/05/2011 2:02:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 02:02:16.765 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:03:00.851 : Writing today.ini, LastUpdateTime = 29/05/2011 2:03:00 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 02:03:08.230 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:02.159 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:05.138 : Writing today.ini, LastUpdateTime = 29/05/2011 2:04:01 AM raindaystart = 7.40000009536743 rain counter = 7.40000009536743
29/05/2011 02:04:08.524 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:11.191 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:13.859 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:16.526 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:19.194 : VP2: LoadCurrentVantageData_V, error = -32701
29/05/2011 02:04:21.862 : VP2: LoadCurrentVantageData_V, error = -32701

Cumulus continued to write the today.ini file at the appointed time, until I closed the program.

[steve]

Yes, so the change I was suggesting would fix (by causing Cumulus to close) the situation where Cumulus continues to log the same data after the DLL has stopped being able to read any new data. You would then be able to (manually) start it up again some time later and it would download the missing data automatically.

There is then a related secondary issue which is the one I asked about a few posts ago; if you had some sort of monitoring software which tried to restart Cumulus again automatically, there is a good chance that whatever the USB problem is would still exist, and Cumulus would just put up a dialogue box and then it would be stuck. Yes?

[Ben]

Yes, Cumulus pops up the dialogue, and needs to be closed manually, so if the usb problem still exists, Cumulus will be waiting for a response from me. This is why I asked if you can catch the problem both at startup and while Cumulus is running.

[steve]

Yes, I could*. It would have to be a hidden option, though, as it would be confusing for it just to close without any indication of what the problem was. Most people wouldn't want it to just close if it had a communication problem at start up (and neither would I, from a support point of view!).

* assuming the auto-close code works in the first place; I have no idea if it does.

[Ben]

I completely understand your thoughts about a support point of view. If you put an option in the configure station window which only pops up when people select Davis and usb, most people would never see it. If you were to tackle it, I could maybe write a new page in the help file explaining what has happened if Cumulus shuts down automatically. Of course, this would be subject to your vetting and editing.

[steve]

I'd just make it an ini-file only setting.

[Ben]

I see the quandry; Cumulus stops for no apparent reason. Your hidden option is probably the best fix then, only those who need to know about it, need know.

It's midnight here and I'm to be up at 6 am, thank you for your time and trouble tonight Steve. I'll be back tomorrow night.

Ben

[pdthull]

Getting the 32701 error trying to connect to VP plus #6161 with firmware Rev B April 24, 2002. It is using a serial connection and weatherlink 5.9.3 shows it connected on Com1 at 9600bps. We shut down weatherlink and try to connect via cumulus on serial com1 and get the error. Does it need to connect at 19,200bps or does that matter?

[steve]

When Cumulus tells the DLL to open the com port, it tells it to use 19200, so I assume it needs to be set to that speed.