I am running everything LoRa (gw, packet forward,influx, grafana, telegraf, redis…) completly on a pi zero w. I am surprised how this little thing is able to handle everything. Thank you Brocaar for your fantastic work.
Currently I do have a esp8266 based measurement infrastructure and am now in the process of switching to LoRaWAN.
I would like my LoRaWAN sensors not to send every minute but to send e.g. every 10 minutes. The measurement on the other hand is happeing every minute. So every 10 minutes a dataframe with 10 measurements is coming in, from which 9 are in the past and 1 is the current one.
I was searching throught the docs and forum but could not find a way to do this directly with the app server. I definetly are able to do this when directly talking to influx, as every insert can have it’s own timestamp (meaning also in the past).
If one could somehow return a special formed object structure from the payload decoder, where timestamps or differential values based on the receive timestamp could be inserted that would be fantastic.
Did you solve this issue? According to https://docs.influxdata.com/influxdb/v1.8/write_protocols/line_protocol_tutorial/ specifying timestamps is possible:
The timestamp for your data point in nanosecond-precision Unix time. The timestamp is optional in line protocol. If you do not specify a timestamp for your data point InfluxDB uses the server’s local nanosecond timestamp in UTC.
I would assume that the InfluxDB Integration is the place to go for this feature.
Hi @John7, a bit late but better than never.
No I did not solve it. I wanted this feature to be an integral part of chirpstack; here the payload decoder. I used external programs to decode the message and write it into influx manually.