Send alerts using data in InfluxDB

Query, analyze, and send alerts using time series data stored in InfluxDB.

This guide uses Python, the InfluxDB v3 Python client library, and the Python Slack SDK to demonstrate how to query data from InfluxDB and send alerts to Slack, but you can use your runtime and alerting platform of choice with any of the available InfluxDB v3 client libraries. Whatever clients and platforms you choose the use, the process is the same:

Alerting process

1. Use an external runtime and InfluxDB client to query data from InfluxDB.
2. Use the queried data and tools available in your runtime to send alerts.

• Create a Slack app
• Install dependencies
• Create an InfluxDB client
• Create a Slack client
• Query InfluxDB
  • Execute the query
• Send alerts
• Full alerting script

Create a Slack app

To send alerts to Slack, first create a Slack app and gather the required connection credentials to interact with your app. More information is provided in the Slack basic app setup documentation.

Install dependencies

Use pip to install the following dependencies:

• influxdb_client_3
• pandas
• slack_sdk

pip install influxdb3-python pandas slack_sdk

Create an InfluxDB client

Use the InfluxDBClient3 function in the influxdb_client_3 module to instantiate an InfluxDB client. Provide the following credentials:

• host: InfluxDB Cloud Serverless region URL (without the protocol)
• org: InfluxDB organization name
• token: InfluxDB API token with read permissions on the bucket you want to query
• database: InfluxDB bucket name

from influxdb_client_3 import InfluxDBClient3
import pandas

# Instantiate an InfluxDBClient3 client configured for your bucket
influxdb = InfluxDBClient3(
    host='cloud2.influxdata.com',
    org='
ORG_NAME
',
    token='
API_TOKEN
',
    database='
BUCKET_NAME
'
)

Create a Slack client

1. Import the WebClient function from the slack.sdk module and the SlackApiError function from the slack_sdk.errors module.

2. Use the WebClient function to instantiate a Slack client. Provide the following credentials:

   • token: Slack bot token

from slack_sdk import WebClient
from slack_sdk.errors import SlackApiError

slack = WebClient(token='
SLACK_BOT_TOKEN
')

Query InfluxDB

Define either a SQL or InfluxQL query to retrieve data to alert on. Depending on what data you want to alert on, you can:

• Include logic in the query so it only returns results that should be alerted on.
• Query data necessary for further processing and then send alerts based on processing performed in your runtime.

The example query below only returns values above a threshold that should trigger alerts.

SELECT
  selector_last(co, time)['time'] AS time,
  selector_last(co, time)['value'] AS co,
  room
FROM home
WHERE co > 10
GROUP BY room

SELECT
  LAST(co) AS co,
  room
FROM home
WHERE co > 10
GROUP BY room

Execute the query

1. Assign the query string to a variable.

2. Use the query method of your instantiated client to query raw data from InfluxDB. Provide the following arguments.

   • query: Query string to execute
   • language: sql or influxql

3. Use the to_pandas method to convert the returned Arrow table to a Pandas DataFrame.

# ...

query = '''
SELECT
  selector_last(co, time)['time'] AS time,
  selector_last(co, time)['value'] AS co,
  room
FROM home
WHERE co > 10
GROUP BY room
'''

table = influxdb_raw.query(query=query, language="sql")
data_frame = table.to_pandas()

# ...

query = '''
SELECT
  LAST(co) AS co,
  room
FROM home
WHERE co > 10
GROUP BY room
'''

table = influxdb_raw.query(query=query, language="influxql")
data_frame = table.to_pandas()

Send alerts

Iterate through the DataFrame and send an alert to Slack for each row.

1. Use the reset_index function on the data frame to ensure indexes align with the number of rows in the DataFrame.

2. Iterate through each row and use the chat_postMessage method of your Slack client to send a message (per row) to Slack. Provide the following arguments:

   • channel: Slack channel to send the alert to.
   • text: Message text to send. Use string interpolation to insert column values from each row into the message text.

# ...

data_frame = data_frame.reset_index()

for index, row in data_frame.iterrows():
    slack.chat_postMessage(
        channel="#
SLACK_CHANNEL
",
        text=f'Carbon monoxide (co) high in {row.room}: {row.co} ppm at {row.time}'
    )

Full alerting script

from influxdb_client_3 import InfluxDBClient3
import pandas
from slack_sdk import WebClient
from slack_sdk.errors import SlackApiError

influxdb = InfluxDBClient3(
    host='cloud2.influxdata.com',
    org='
ORG_NAME
',
    token='
API_TOKEN
',
    database='
BUCKET_NAME
'
)

slack = WebClient(token='
SLACK_BOT_TOKEN
')

query = '''
SELECT
  selector_last(co, time)['time'] AS time,
  selector_last(co, time)['value'] AS co,
  room
FROM home
WHERE co > 10
GROUP BY room
'''

table = influxdb_raw.query(query=query, language="sql")
data_frame = table.to_pandas()
data_frame = data_frame.reset_index()

for index, row in data_frame.iterrows():
    slack.chat_postMessage(
        channel="#
SLACK_CHANNEL
",
        text=f'Carbon monoxide (co) high in {row.room}: {row.co} ppm at {row.time}'
    )

from influxdb_client_3 import InfluxDBClient3
import pandas
from slack_sdk import WebClient
from slack_sdk.errors import SlackApiError

influxdb = InfluxDBClient3(
    host='cloud2.influxdata.com',
    org='
ORG_NAME
',
    token='
API_TOKEN
',
    database='
BUCKET_NAME
'
)

slack = WebClient(token='
SLACK_BOT_TOKEN
')

query = '''
SELECT
  LAST(co) AS co,
  room
FROM home
WHERE co > 10
GROUP BY room
'''

table = influxdb_raw.query(query=query, language="influxql")
data_frame = table.to_pandas()
data_frame = data_frame.reset_index()

for index, row in data_frame.iterrows():
    slack.chat_postMessage(
        channel="#
SLACK_CHANNEL
",
        text=f'Carbon monoxide (co) high in {row.room}: {row.co} ppm at {row.time}'
    )