Influx Alert multiple fields configuration

InfluxDB 2 Checks & Notifications
1

Hi guys,
I am a new developer and also new to influxDB. I am working on alerts/checks and notifications, after following the steps from the document, I am able to see the alert from InfluxDB UI. However, I realised that I could only select ONE field to check for ONE sensor. What should I do if I have multiple sensors and fields that need to be checked? How can I configure it?
Thank you for reading.

2

Hello @vinh_lee,
So I don’t recommend using the UI to make alerts. Instead use tasks.
This way you can use any custom logic you want using Flux:

import "array"
import "slack"

option task = { name: "Event Alert", every: 1h0m0s, offset: 5m0s }

alert = (eventValue, threshold) =>
   (if eventValue >= threshold then slack.message(
       url: "https://hooks.slack.com/services/####/####/####",
       text: "An alert event has occurred! The number of field values= \"${string(v: eventValue)}\ ". and The sensor is= \"${string(v: sensorName)}\ ". ",
       color: "warning",
   ) else 0)
data = from(bucket: "bucket1")  
   |> range(start: -task.every, stop: now())
   |> filter(fn: (r) =>
       (r._measurement == "measurement1" and r._field == "field1" or r._field == "field2))
    |> group()
    |> filter(fn: (r) => r._value >  threshold, onEmpty: "keep")

events = data
   |> findRecord(fn: (key) =>
       (true), idx: 0)
eventTotal = events._value
sensorName = events.<SensorTagKey>

data_0
   |> yield(name: "ignore")
alert(eventValue: eventTotal, threshold: <your threshold value>)

Alternatively I really recommend using Grafana for alerting lol.

something like that though.

PS what are you doing with influxDB? I like to learn about community users use cases. Thank you!

1 Like
3

Hi @Anaisdg. Thank you so much for your reply, It is just for my team project when we want to check our sensor. I already solved the issue using custom tasks as you suggested. However, I have another question about custom tasks or notifications. I wonder what happens if I intentionally
set up “every” one second for my task to show the error immediately. Do you reckon it is a good practice? if not, how do we send the message?

Thanks

4

Setting it up to run every second really depends on how much data you’re querying each time. It’s not necessarily a bad idea if you’re querying a small amount of data.
However you could also consider using one task for the processing where you then write the alert status to a new measurement or bucket and a second task for the alerting that runs every second. The second task would query that new measurement and bucket and return an alert if there is a value present.

1 Like
5

Hi @Anaisdg ,
Thank you so much for your response,
Since my sensor will write its measurement every one second to a bucket, I’m not going to query big data, probably within -10 or 15 seconds.

However, to check its status, I will need a task, which will run every one second, to store the status. Do you think it duplicates the data, for instance, 1 bucket for real database and another bucket just for status ? Let’s say in 24 hours, there are no changes in the status.

Apologize for asking so many things. I really want to know your point of view. Not only InfluxDb but also the Database concept.

Regards,

6

@vinh_lee,
Of course
Yah I’d probably put the results of a check in another bucket with a short retention policy so you’re not keeping that data around forever. Unless of course you want it for some reason.

1 Like
7

Thank you so much. :smiley: