Kubernetes compatible dashboards

[blackjid]

Hi,

I've been working on running lnd on kubernetes. Using lndmon with grafana/prometheus is awsome for visibility.

However, the provided dashboards don't work that well if you run multiple nodes on the same cluster because you don't have a way to filter by node and all the metrics are mixed together.

This PR adds some variables to filter the metrics by:

• datasource: this is useful if you have multiple nodes scraped by different prometheus instances.
• namespace: to filter by namespace :)
• job: to filter by job.

I also added the lightning-network tag for the dashboards. (this may have been done in a separated pr)

These branch dashboards are used in a helm chart I wrote but I would be better if we can use the "official" dashboard provided in this repo. https://github.com/budacom/helm-lnd

Note

I'm not really sure how these changes affect the use of the docker-compose deployment proposed in the repo. That is why I'm marking this PR as draf, it would be asewome to iterate on the changes to make them compatible, of find a way to provide dashboards of all kind of deployments.
I have made some changes so it works both in the proposed deployment and in kubernetes using default labels namespace and job

[mrfelton]

Have been using this in production for several months. Does the job. Would be awesome to see this updated and merged

[carlaKC]

Haven't got a local setup to test this out, but tested that this doesn't adversely affect the non-kubernetes setup and all good in that regard ✅

@lispmeister can you do a quick skim of this as well? Can confirm that it doesn't break anything, and I take the honorable @mrfelton at his word that it's working.

[mrfelton]

Looking back over this patch I see that I actually made some changes to it for our setup. Rather than adding $namespace and $job tags, in my version I used $namespace and $node which worked out to be a more useful grouping of things. It's such a long time ago that I forked the PR that I honestly can't remember the specifics of why I changed it that way.

@blackjid do you have any thoughts about this?

[mrfelton]

Here is the version that we are running in production: #60

[blackjid]

#60 looks perfect for me... 💯

[lightninglabs-deploy]

@lispmeister: review reminder

[guggero]

Replaced by #60.