Missing Data in Logfile after a shutdown

[KD7YZ]

I shutdown the Cumulus program around 2330 last night. I restarted it around 0715.

Later, I wanted to see what the recorded outside temperature at 0600 was.

So I tried reading a file which seems to go from Mar-15 to present. I did this by reading from the monthly data log viewer seen in my file upload.

However, the data for the time period is missing.

When I'd started Cumulus, I'd seen the Pop-Up stating that the Davis was being read from.

[steve]

Please zip up the diags folder and attach it to a reply in this thread.

[KD7YZ]

here is zip file

thanks for the help. Sorry about being such a Newbie

[steve]

When you started it up (apparently at 0515), it was initially unable to connect to the station, so it couldn't download the required archive data. Shortly afterwards, it was able to connect, so it was able to read the live data. The system had apparently only been up for 16 seconds when Cumulus started, so it seems likely that the network connection wasn't yet available. It looks like you're starting Cumulus automatically when the PC starts up; you may need to do something to make sure the connection is available before Cumulus is allowed to start.

If you stop Cumulus and find the backup folder that it created when it started up:

30-Mar-15 05:15:53.784 : Creating backup folder C:\Cumulus\backup\20150330051548\

and copy all of the files from that folder into your data folder, overwriting the ones that are already there, and then start Cumulus up again, it should download the required data and you should no longer have a gap.

Possibly related to this, your PC clock seems to have done something odd just after Cumulus started, and just at the time it was trying to connect to the station:

30-Mar-15 05:15:55.324 : VP2: CloseCommPort_V, res = 0
30-Mar-15 05:15:55.324 : VP2: Open connection, type = 2
30-Mar-15 05:15:57.327 : VP2: OpenTCPIPPort_V, res = -32701
30-Mar-15 07:14:24.170 : Drawing initial graphs from data logs
30-Mar-15 07:14:25.000 : Finished drawing initial graphs from data logs
30-Mar-15 07:15:25.693 : begin updating station settings

The clock suddenly jumped forwards by two hours.

[KD7YZ]

Starting the computer this morning, thus Cumulus, I immediately noticed that the high temp yesterday's period (0600 to 0600) for my NOAA report, was 100 degrees and the low was -18. As seen on the face of Cumulus.

I replaced the files suggested.
Then a series of events started after I restarted Cumulus.
If the uploaded files remain in the order 1 ..5, then is will be more easily understood.
First nothing happened for several minutes as I waited for the 'downloading data' to finish.
It did not, so I killed the process-tree with sysinternals process explorer.

I restarted Cumulus. Saw error light and read about, among many new errors, that there was now a duplicate entry.
There was. I deleted the entry. I tried to restart but Cumulus only grayed out parts oif the GUI.

I killed the process again.
Went over to the Davis program to see if comms with the Davis VP2 were possible. They were as seen by a snapshot of VP2 download of 513 records.

I'm not sure of the sequence position, in the timeline of the BSOD which occurred ... might have been one up from this last line or the Next" position ... but there was a hard BSOD while Cumulus was processing the newly overlayed files from the backup folder into the present folder, per your instructions in your last posting.

The sys restart and then Cumulus went off just fine. And the final (05) jpeg shows the GUI status plus the still-present errors minus the dupe-line 855 which, as I wrote, I'd deleted in Dayfile editor.

[steve]

The error in the March log file is likely caused by the crash (BSOD), or by you killing the process. The error should be obvious in a good text editor. A BSOD is usually caused by a hardware or driver problem.

Can you explain what you mean by

but there was a hard BSOD while Cumulus was processing the newly overlayed files from the backup folder into the present folder, per your instructions in your last posting.

I don't understand what you mean by "while Cumulus was processing". It doesn't process any existing data logs. Do you mean while it was downloading data from the logger, when it started up after you had replaced the files from the backup?

Please zip up the diags folder and attach it.

[KD7YZ]

The error in the March log file is likely caused by the crash (BSOD), or by you killing the process. The error should be obvious in a good text editor. A BSOD is usually caused by a hardware or driver problem.

regards me killing the non-responsive task, well, I saw no other way to gain access to the GUI anymore, at that point.
This spate of non-responsiveness started when Cumulus stalled while trying to read the archive from the VP2. I tried the "X" to stop it but there was no response.

That may be, or quite possibly IS the case. That said, in the 4 or 5 BSOD in the last few cumulative hours while Cumulus is running, there was not a blue-screen-of-death whilst Cumulus was NOT running.

Now, today, on that computer, there were two.

I am wondering if the, still persistent, error messages about the 'index out of bounds' could cause that particular computer to crash?
How do I determine which line is ["line 32029" -1] in "Mar15log.txt" ? So I can attempt a repair?

Barring that, I suppose I could wipe it all clean and re-install, re-import wd-data, and so forth, tedious as that may be.

The enclosed jpg is current.

thanks for thee continuing help.

[steve]

I am wondering if the, still persistent, error messages about the 'index out of bounds' could cause that particular computer to crash?

Highly unlikely. As you have an IP logger, it is much likely to be a problem with the driver for your network card, or perhaps a disk problem, perhaps brought to the fore by the particular way that Cumulus uses them. If you quote the details from the BSOD it may be possible to say where the problem lies. There's some good information about BSODs here - http://www.howtogeek.com/163452/everyth ... -of-death/

How do I determine which line is ["line 32029" -1] in "Mar15log.txt" ? So I can attempt a repair?

In notepad you can use 'go to line number' on the edit menu, or (better) use a good text editor like notepad++ which will show the line numbers and also make some kinds of file corruption more obvious. If you go to the notepad++ web site to download it today, bear in mind what the date is, the site isn't usually like that!

I'm still happy to look at the diags files if you upload them as I asked.

[KD7YZ]

Highly unlikely. As you have an IP logger, ........ If you quote the details from the BSOD it may be possible to say where the problem lies. There's some good information about BSODs here -

Ahh, perfect. I will peruse them and hopefully have a line or two to post back about the BSOD's ...meanwhile ...

....or (better) use a good text editor like notepad++ which will show the line numbers and also make some kinds of file corruption more obvious. If you go to the notepad++ web site to download it today, bear in mind what the date is, the site isn't usually like that! ...

OMFG, WTF .... a Chameleon rock band?

Oy! next I'll read MU beats FC-Bayern

I D/L Notepad++, enduring the Chameleon and large icon.
noted that the referenced line error were all NUL.

Deleted that line. Closed correctly. Restarted, waited appropriate time for VP2 download. Checked Errors and that file is essentially blank now. Problem ostensibly solved.

I'm still happy to look at the diags files if you upload them as I asked.

yikes! my bad, I forgot.

Again, thanks.

[steve]

The diags for yesterday morning are gone now, Cumulus has been started at least ten times since then (the logs for the last 10 starts are kept). I see that you have the logger interval set to one minute, so it's possible that in the first run yesterday which was taking a long time to start up, it was simply having to process a lot of logger entries.

There's nothing in there to give any clues as to the BSODs, but then I wouldn't expect that.

[KD7YZ]

The diags for yesterday morning are gone now, Cumulus has been started at least ten times since then ...
.....There's nothing in there to give any clues as to the BSODs, but then I wouldn't expect that.

arguably true.

re the 10 restarts, on this computer, for the time being ....due to the BSOD's. Thus I didn't actually notice (each of) the restarts.

Will leave all alone and just see how it does for a few days.

I _do_ have issues, only on this computer, with instances of shockwave-flash calls from, usually, Firefox. Many times that alone makes BSOD. Can't trace the problem yet. Later I'll use the BSOD links supplied by you.

SO, how does normally running Cumulus react to a hard-system-crash and subsequent auto-restart?

[steve]

SO, how does normally running Cumulus react to a hard-system-crash and subsequent auto-restart?

It just picks up from where it left off - from the last time it wrote to the log files. But of course the crash may have caused file corruption.