New alerts are not taking effect

Help
1

I have Netdata 1.45 running in a Docker on my units. I have followed the documentation to the ebs too my ability however a test alert I am setting up to alert me when the bandwidth usage of our system goes over a certain amount does not appear to be working. Perhaps someone can shed some light?

I did the following:

-touch health.d/custom.conf
-./edit-config health.d/custom.conf

Pasted in:

template: mainheads|system_net|1_3512eb07-ff49-4d1b-9e3a-9abfab591873
on: mainheads|system.net|1
lookup: average -5m
every: 60s
crit: $this > (85)
warn: $this > (.015)
info: Test
to: sysadmin

Then I did a

netdatacli reload-health

and I still don’t see an alert. I imagine I have set this low enough to see an alert with pretty much any traffic. Also, it’s important to note that I got most of the alert text from the new alert config GUI by clicking the alarm symbol next to the metric I was interested in.

Other alerts are present on the system (presumably from the default set) however this one is not added.

2

@lfstudios10 : I see a couple of issues with your configuration:

  1. on: This defines the context on which you want to create the alert. And afaik, the network interfaces bandwidth metric context is system.net. I am not sure if mainheads|system.net|1 is a valid context.

  2. Units: The default unit of the alert comes from the context and for the system.net context, the units are gigabits/s and not percentage.
    So, the warning / critical alert should be triggered purely based on the physical bandwidth consumption of the network interface.

  3. You are creating this alert based on all the dimensions (sent / received) of the system.net chart. Which basically means, you are checking for the average of (sum of received + sent bandwidth) over the last 5 minutes. You may want to check if you want to create separate alerts for sent and receive using the foreach construct.

I am sharing the alert configurations reference guide with you for more details - Configure alerts | Learn Netdata

Hope this helps!

3

  1. This is what was provided by the alert configurator otherwise I have no clue how I would have generated it.

  2. While I understand what you are saying, can you clarify exactly the syntax that needs ot change

  3. I used the linked guide to get as far as I did but I am definitely not doing things properly. For each would have one alert for sent going above a certain amount and one for receive?