This script is currently written to explictly log data for just the last 1 minute assuming that your cron/scheduler is the throttle for data fidelity. There's lots of options available from flume. Feel free to make this code more robust/configurable

• python3
• pyjwt library
• requests library

• Your API client ID and secret. You have to get this through the flumetech portal: https://portal.flumetech.com/#dashboard
• Your username and password for the portal.
• Only 1 flume device. I wrote this code to work with a single flume sensor. I have no idea what will happen if you have more than one.
• If you want to send to Splunk you will need to log to a file (using --logfile) and collect with a UF, or you need an HTTP event collector token/URL.
  • I have no idea if this works with HTTPS HEC. Try it out :D

1. First you need to establish a JWT/Token. This requires client ID & secret as well as username and password. You need to specify a tokenfile to write the resulting token to.
   • flumecli.py --auth --clientid <clientid> --clientsecret <clientsecret> --username <flumetech username> --password <flumetech password> --tokenfile <pathtofile>
2. You may want to have a look at the details of your environment. These can be used to interface with the API directly or using other systems.
   • flumecli.py --details --tokenfile <pathtofile>
   • Your User ID and Device ID are generated by the flume system. These are often required when interacting with other API calls
3. Query the flume API. There's a query language from flume but for the purposes of this script I'm just looking at the last 1 minute of water flow, assuming that you just schedule this script to run every minute. There's a number of different ways to output this data.
   1. flumecli.py --query --tokenfile <pathtofile> Simple query with output to stdout showing timestamp and water flow from last minute
   2. flumecli.py --query --tokenfile <pathtofile> --logfile <pathtologfile> Same output as above, except the output gets appended to the specified file
   3. flumecli.py --query --tokenfile <pathtofile> --hecurl <full hec url> --hectoken <hec token> Outputs the value directly to HEC