@Eucalyptus: If you are not familiar how to generate exports. I recommend to use the NOAA iso template, like below.
Bildschirmfoto 2022-07-04 um 11.02.32.png (39.15 KiB) Viewed 750 times
Please generate in resolutions "hour" and "day" and for period "202206010000" to "202207010000" to get full June exported. "Rain" column shows logged and stored rain data. If you compare that to the console values it should match. BTW, you find exported files in SMB share of your Meteobridge, where any windows PC can connect to as a pc-network drive. Or you can call "http://ip-of-your-meteobridge/cgi-bin/download.cgi" where you can grab files from "export" folder as well.
Rainfall comparison update now the rain event is basically over and how the 5 systems tracked. One Nano and one MBPro were rebooted during the rain event. Firmware remains unchanged with previous comparison
Of interest was the all the monthly totals agreed exactly yet one MBPro on registered 10.2mm today and the Envoy/WLIP 10.8mm and both these dependencies are time related due to constant rain before, during and after midnight. The MBPro had been up for 496 days and the time had slipped by 22 minutes, rebooting fixed the time but not the rainfall which had already been timestamped. Similar with the Envoy/WLIP time off by a couple of minutes hence the 10.8mm. Nothing in that is any big deal but obviously time maintenance is something to keep in mind over long period of continual up time.
NanoSD61
Nano63
MBPro75
MBPro78
Davis Envoy/WLIP Monthly total 146.6mm, Rain Today 10.8mm
Rainfall comparison update now the rain event is basically over and how the 5 systems tracked. One Nano and one MBPro were rebooted during the rain event. Firmware remains unchanged with previous comparison
Of interest was all the monthly totals agreed exactly yet one MBPro only registered 10.2mm today and the Envoy/WLIP 10.8mm and both these discrepancies are time related due to constant rain before, during and after midnight. The MBPro had been up for 496 days and the time had slipped by 22 minutes, rebooting fixed the time but not the rainfall which had already been timestamped. Similar with the Envoy/WLIP time off by a couple of minutes hence the 10.8mm. Nothing in that is any big deal but obviously time maintenance is something to keep in mind over long period of continual up time.
NanoSD61
Nano63
MBPro75
MBPro78
Davis Envoy/WLIP Monthly total 146.6mm, Rain Today 10.8mm
Further debugging isolated the root cause. Meteobridge internal computation assumes that defined timezone do not have a sub-hour difference to GMT. Affected Meteobridges are set to timezone "Adelaide South Australia", which is GMT+9:30. GMT+9 or GMT+10 works fine, but GMT+9:30 is generating problems.
Meteobridge is internally working with epochtime (seconds til 1970-01-01 00:00:00) and assumes that change of hour can be identified by doing (epochtime % 3600). This is not right when timezone has a sub hour offset. It will take some time to find and update all the impacted code segments. For the moment I recommend to avoid time zone settings with sub hour GMT offsets.
