Losing live WMR200 data at start-up

Since recent changes to avoid losing data when flat-lining, I now seem to be losing live data every morning at start-up for the time it takes to process the overnight WMR200 history data, typically 7-9 minutes.

E.g. I start WD at 0930, the graphs are updated by 0937, the first live data is received at 0938, and I’m missing data for 8 minutes (0933 to 0940 inclusive) in the daily log.

replying here to your same post on your other thread:
I have not changed any code around processing the live data during the history extraction
(so not sure why this is happening)

Nor am I, but superficially it seemed similar to the recently modified behaviour for flat-lining - no data, no logging, DQ light goes red - but in reverse, i.e. if DQ is red, no logging. . .

E.g. this morning I started up at 0929, and both Data lights were red. DQ went green at 0933, but DR didn’t go green until 0935. Log entries missing for 0929-36.

Any thoughts yet as to why this is happening?

I attach an extract from the log for 5 April, as mentioned above, with 0930-36 missing. . . I see 0929 is actually present, but out of chronological order like all the overnight history data.


Log.JPG

Any ideas yet, Brian?

I had hoped that the longer “stop logging” threshold in build 46 would cure this problem, but it hasn’t. This morning I started WD at 0931, DQ light went green at 0935, overnight graphs were updated at 0936, DR went green at 093630, data values appeared above graphs at 0937, but I’ve lost 7 minutes (0932-38) in the log.

Now running the latest 10.37S47 (24 April) but I’m still losing 5-7 minutes of live data every morning while the overnight history data is being processed.

I note that you haven’t changed any code around processing live data, so you are not sure why this is happening, but you have changed the conditions that govern logging. Please can you have another think about this, it’s driving me nuts.

WD should be picking up the live data that arrives while the history data comes through and then should process that at the end of processing the history data

at least that is how it worked in the older build 81

are you not able to leave WD running overnight?

You changed the history data collection at build R320 last March so WMR200 History vers 1.3 gathers all the history data and then hands over to WMR200 History Data to decode it, so build 81 doesn’t come into it.

[Handover takes some time, and wmr200history.txt now logs raw data rather than decoded data, so not much use to me (see separate post).]

But the current problem, losing live data, has only happened since the modification to stop logging during flat-lining.

And yes, I could leave WD running overnight but I choose not to. With respect, why should I have to leave it running?

For clarification, I am of a generation that turns off everything non-essential at night: computer, TV, tablet, mobile phone, battery chargers, etc. I don’t go quite as far as my parents did and unplug everything, though!

But I bought a weather station with a built-in data logger so I didn’t have to leave the computer on overnight.

Any ideas yet, Brian?

Apologies: as described elsewhere I have discovered that this continuing problem is nothing to do with the recent flat-lining mod, by reverting to an earlier build for a day.

Unfortunately, that means that it’s been happening unnoticed for a while.

While I digest my humble pie I intend to go back in the logs as far as necessary to see if I can put a start date on it.

Did I say digest? Indigestion, more like. . .

I can say that the WMR200 “live data loss at start-up” problem has been occurring since I installed 10.37R320 on 10 March 2016 (here).

I think it was also happening before this, but my logs were horribly fragmented before R319 (4 March 2016, loc.cit.) and still confused in the week between R319 and R320. . .

After discussion elsewhere (for a different problem) about USB-over-ethernet cable I experimented at start-up this morning with a direct USB connection. As expected, this made no difference whatsoever: I started WD at 1031, DQ light went green at 1035, overnight graphs were updated at 1036, DR went green at 1036:30, data values appeared above graphs at 1037, and I’ve lost 7 minutes (1031-37) in the log (see attachment). Note the chronological confusion of the last few history data entries.


dailylog.JPG

It would be interesting to see if you can replicate this, Brian, or point to a problem with my settings.

Just occurred to me: would I not still have the same problem when restarting after a power outage?

And I get the same problem when restarting after flat-lining, like today.

a UPS would help re power outage

I am not getting other reports of the flat lining problem you are getting
which points to your hardware and something about that or your windows (and power settings)

This thread has nothing to do with flat-lining or UPSs, it is to do with the fact that WD doesn’t log live WMR200 data when it’s busy processing history data from the console logger. I’ve been away for a few days and I just lost 26 minutes of live data when I restarted.