ohAnd
Interruptions Data Hoymile Cloud
I see some interruptions in the data of the Hoymile Cloud (value goes to zero) because value can't delivered or is blocked from DTU-Gateway. Happens differently, some day only 1-2, sometimes 3 zero values. Router is Fritzbox 7530AX.
A fresh poweroff in the morning seems not changing behavior really.
What I see sometimes a bigger unsynchronity between time gw local and gw ntp (>20-30s). Is this a problem or maybe the reason? Could a change of the value from dtu request cycle help ( currently default 31s) ? Could a reboot (time configurable) in the night after inverter is switched off maybe help?
Could it be that the ESP-board is not running stable/ fine with clock because of big time differnces between gw local and gw ntp ? Timing of clock instable or something like this ?
Remark: I have another dtu-gateway in operation on another location in combination with DTU-Wlite and HMS-800 ( is also working ) there I don't see interruptions in the cloud data, only very less (~1 in 15 days). But also running with other router combination and other ESP-board.
attachments:
On that day I have made a fresh poweroff in the morning, see gw start time.
cloud data from day 2.6 DTUGateway status 2.6 evening DTUGateway status 3.6 night/morning
basically it could be the cause if the gw local time has such a big diff to the ntp time... based on this time the cloud pause will be given, to give the dtu the possibilty to send data to the cloud (and is so not "disturbed" by local requests)
What is the gw local time: This time is a local clock which will be constantly synced with the clock of the dtu. Means with every request the dtu delivers also a timestamp in their response this will be used to synchronize. So the diff is inner the dtu between ntp time and local time (and so also forwarded in dtuGateway -> local time)
So, here could be the question - in comparision to your working sample - is the overall connection between dtuGateway / dtu/ local wifi weak, so there are maybe not every request or lot's of them couldn't be processed?
And with this, dtu has 'unfortunately' to this points in time no connection to the hoymiles servers (to upload data) and maybe the same point in time will be used to synchronize the clock of dtu with ntp time externally... this can cause in the time diff of dtu (and so also for dtuGatewas localtime) in comparison the overall ntp time.
has to be supervised regarding the wifi connection
as you already mentioned reboot of dtuGateway gives no change - and from my perspective it will not change anything with the given malfunction behavior
The connection should be normally be fine. As it is only 6m distance between router and inverter. Also cloud-data is always existend when I switch off dtuGateway, so connection should be ok.
So only the timestamp is synchronized ? But not the time itself? Makes it sense to have the possibility via switch in advanced preferences to synch ntp-time with local-time ?
Unfortunatly I must switch off dtuGateway currently because since the last FW-update from DTU which increases the data in the cloud to 5min time-slot, it seems it's in generell not more working. So we need a possibility to adjust the parameter which is currently 15min to a flexible parameter to change it depending which FW is active in the inverter.