Cumulus Support

MX 3.0 (3036) on GNU/Linux (Ubuntu 15.10) attached to VP2 via Virtual VP
I looked at the monthly records page today, January in particular, and noted that the rainfall was incorrect. The figure for December 2015 appears. See http://www.hosiene.co.uk/weather/monthlyrecord.htm. I only have recorded one January (2015) and its rainfall total is the one I expected to see displayed.

This presumably happened around the end of year/month rollover. It may be that it gets the month wrong for 0900 rollovers for this item (the first nine hours of January should count to December, as you will know). Do you still have the diags files from then?

I've looked at the code and it seems to be correct. I'm wondering if it's possible that a reading came in at or just after 0900, but just before the rollover had been done, so the daily total hadn't been reset. The diags file would prove this.

steve wrote:This presumably happened around the end of year/month rollover. It may be that it gets the month wrong for 0900 rollovers for this item (the first nine hours of January should count to December, as you will know). Do you still have the diags files from then?

I'm afraid I haven't, sorry.

steve wrote:I've looked at the code and it seems to be correct. I'm wondering if it's possible that a reading came in at or just after 0900, but just before the rollover had been done, so the daily total hadn't been reset. The diags file would prove this.

I don't have the diags file, but I do have the monthly data files. There was a rain tip at 17:33 on 31/12/2015. the next one was at 14:04 on 01/01/2016.
Is there anything else in the files that you would like me to look at?

Thanks, but no - only the diags file would have confirmed my theory. It doesn't matter when the actual tips were, the rain gets processed every time the station sends data.