Tutorial Developer SensorClientFIFOWrite
This tutorial describes how to write your own scripts that writes into the FIFO file of the AlertR Sensor Client FIFO. This tutorial assumes you have set up the AlertR Sensor Client FIFO according to the provided AlertR Sensor Client FIFO Tutorial and only describes the necessary additions.
The following gives you a short example on how to configure the AlertR Sensor Client FIFO and write a script that writes processable data for AlertR. The script is written in Python, but it can be any programming language you like. I just find Python most convenient to show what has to be done.
The corresponding sensor configuration for the AlertR Sensor Client FIFO looks like the following:
[...]
<sensors>
<sensor>
<general
id="0"
description="Developer Sensor"
alertDelay="0"
triggerAlert="True"
triggerAlertNormal="True" />
<alertLevel>0</alertLevel>
<fifo
umask="0000"
fifoFile="/home/alertr/sensorClientFIFO/some_notification.fifo"
dataType="1" />
</sensor>
</sensors>
[...]
This sensor has a FIFO file at /home/alertr/sensorClientFIFO/some_notification.fifo and the data that is provided by the sensor is of the type integer (1 = integer).
The following shows the script that writes into the FIFO file. This script does not have any purpose at all. It should just demonstrate how to output data that can be processed by AlertR.
#!/usr/bin/python3
import json
import random
import time
# This function creates a Sensor Alert message dictionary.
# It uses the given "value" as data for the sensor alert,
# the "msg" as message for the optional data,
# and the "state" as state for the sensor alert.
def create_sensor_alert(value, msg, state):
result = dict()
result["message"] = "sensoralert"
optionalData = dict()
optionalData["message"] = msg
payload = dict()
payload["state"] = state
payload["hasOptionalData"] = True
payload["optionalData"] = optionalData
payload["dataType"] = 1
payload["data"] = {"value": value, "unit": ""}
payload["hasLatestData"] = True
payload["changeState"] = True
result["payload"] = payload
return result
# This function creates a State Change message dictionary.
# It uses the given "value" as data for the state change,
# and the "state" as state for the state change.
def create_state_change(value, state):
result = dict()
result["message"] = "statechange"
payload = dict()
payload["state"] = state
payload["dataType"] = 1
payload["data"] = {"value": value, "unit": ""}
result["payload"] = payload
return result
# This actually does nothing. It just creates a
# random number and triggers a sensor alert if
# it is greater than 10. The created number is used
# as data for the message.
def main():
fifo_file = "/home/alertr/sensorClientFIFO/some_notification.fifo"
while True:
data = random.randint(0, 20)
msg = None
if data > 10:
msg = create_sensor_alert(data, "Oh no, over 10!", 1)
else:
msg = create_state_change(data, 0)
with open(fifo_file, "w") as fp:
fp.write(json.dumps(msg))
time.sleep(5)
if __name__ == '__main__':
main()Done. This script creates a sensor alert if a random number is greater than 10. Otherwise, a state change message with the state normal is returned. With a FIFO file you can trigger an event in the AlertR system as soon as it occurs (in contrast to the AlertR Sensor Client Executer which executes scripts in an interval). If you want a detailed description of what messages can be passed back to the AlertR system via writing to a FIFO file, please take a look at the Protocol of the AlertR Sensor Client FIFO.
If you want additional examples that have actually any purpose, please take a look into the scripts_example folder in the AlertR Sensor Client FIFO installation directory.
If you experience problems, please check the log file first. If it is not helpful, change the log level to DEBUG and check again. If no error can be seen, please start the AlertR client manually and check if an error occurs that is not printed into the log file. This can be done by just executing the AlertR client as the user that it normally runs with.
alertr@towel:~/sensorClientFIFO$ ./alertRclient.py
If you still have problems and do not know how to solve them, you can ask on the community page on reddit or you can use the Github Issues.