HTTP JSON Sensor Data

Aus Gude Systems GmbH
Zur Navigation springen Zur Suche springen

Preface

  • all Sensor data is available in a generic JSON Object, presentes by status.json
  • there are two relevant sub-objects: sensor_values and sensor_descr
  • sensor_descr can tell:
    • what sensor types are actually relevant
    • what is the type-specific sensor-field description
    • how many sensors of each type are actually present
    • what are the names of each of those sensors
  • sensor_values tells you about the sensor values, matching the description explained above
  • it's common practice to get sensor_descr+sensor_values once, and poll sensor_values only subsequently with mich less payload

This might look complicated at first glance, but this empowers you to write small code supporting all Products and all Sensors in one single approach, without the need for sensor specific knowlege. So your software is already prepared for new Gude-Devices, and prepared for new Sensor Add-Ons. As example, this documentation will highly depend on check_gude.py, our HTTP sensor data swiss-knife tool.


Getting Data

  • HTTP-Get status.json?components=8470528
    • here, 8470528 sums up sensor_values (16384 aka 0x10000) plus sensor_descr (65536 aka 0x4000), plus the ‘extended’ marker (0x800000) to get both simple sensors and complex sensor groups
    • 0x4000 + 0x10000 + 0x800000 = 8470528

Example Data

sensor_descr

 [
   {
     "type": 664,
     "num": 2,
     "fields": [
        {"name": "Voltage", "unit": "V", "decPrecision": 3},
        {"name": "Current", "unit": "A", "decPrecision": 1}
     ]
     "properties": [
       { "id": "L1", "name": "Meter1", "state": 1},
       { "id": "L2", "name": "Meter2", "state": 1}
     ]
   },
   {
     "type": 665,
     "num": 1,
     "fields": [
        {"name": "Temperature", "unit": "C", "decPrecision": 1},
        {"name": "Humidity",    "unit": "%", "decPrecision": 1}
     ]
     "properties": [
       { "id": "6102", "name": "Server-Rack", "state": 1}
     ]
   }
 ]

This tells you:

  • there a two Sensors of Type 664, and one of type 665
  • a type-664 sensor has two Fields, Voltage and Ampere
    • Voltage is measured with a decimal precision of 3, Ampere with a decimal precision of 1
  • The two Type-664 sensors (L1 and L2) are named 'Meter1' and 'Meter2'
  • The Type-665 Sensor is named 'Sever-Rack'

sensor_values

 [
   {
     "type": 664,
     "num": 2,
     "values": [
       [{"v": 233.19}, {"v": 3.2}],
       [{"v": 226.2},  {"v": 0.3}]
     ]
   },
   {
     "type": 665,
     "num": 1,
     "values": [
       [{"v": 27.1}, {"v": 40.3}]
     ]
   }
 ]

sensor_desc / sensor_values

bringing together this tells you:

 L1/Meter1 233.19 V (Voltage), 3.2 A (Ampere)
 L2/Meter2 226.20 V (Voltage), 0.3 A (Ampere)
 
 6102/Server-Rack: 27.1 C (Temperature), 40.3 % (Humidity)

Common Sensor Type IDs

  1  Line power meter
  9  Line power meter with residual Current
  7  Digital Inputs
  8  Outlet power meter
 20  System Data (sensor group)
 51  Temperature Sensor
 52  Temperature/Humidity Sensor
 53  Temperature/Humidity/AirPressure Sensor
101  Bank (eFuses Port-groups) Sensor (e.g. used at 8291 PDUs)
102  (DC) Power Sources

check_gude.py in action

  • check_gude.py is a demo code to show how sensor_descr and sensor_values can be assembled generically to make use of all out devices / sensors
  • so when new devices and sensors are coming up, check_gude.py is already prepared to deal with it
  • install python along with requests to run check_gude.py
  • feel free to use check_gude.py as you need it and to rewrite the given code to any language desired

Show all Sensor Data

Here a device with hostname 8041.demo.gude.info is queried to dump all sensor data

Check gude py 1.png

show CGI-Get / JSON Data

when using --verbose check_gude.py will print out the full URL and the JSON return data:

Check gude py 2.png

Sensor Groups