Support for custom annotations in Helm Chart

Issues Dgraph
1

Experience Report for Feature Request

As an operator of dgraph, I’d like to monitor the system using the tools and technologies that my company currently supports. The dgraph helm chart comes prepackaged with prometheus annotations. However, some companies use solutions like Datadog as their defacto monitoring solution. Allowing consumers to specify custom annotations will allow them to leverage their monitoring tool of choice.

What you wanted to do

Add a podAnnotations field to the alpha and zero configuration blocks to allow consumers to plug in their own custom annotations.

What you actually did

I have not yet implemented this, but it appears to be rather straightforward to do.

Why that wasn’t great, with examples

N/A

Any external references to support your case

N/A

2

From some of my digging…

It looks like Datadog’s agent can be configured to use the prometheus annotations: https://docs.datadoghq.com/agent/kubernetes/prometheus/

It would still be useful to use the autodiscovery annotations for some processes rather than opting the entire cluster into it.

3

We’ve added config support for alpha.extraAnnotations, zero.extraAnnotations, and ratel.extraAnnotations in this PR:

I hope that fits your use case @mjpitz. We’ll release this new chart version publicly this week.

1 Like
4

@mjpitz The v0.0.18 Helm chart has been published. You can update to the latest helm chart to use the extraAnnotations config. I hope that helps with your monitoring setup.

5

Thanks for following up! I got this all going. Thanks :slight_smile: