Move our Grafana dashboards to spacemeridian.grafana.net

[bajtos]

Public Dashboards

• Spark Public Dashboard
• Station Public Dashboard

Out of scope:

• Custom domain for Station dashboard: dash.filstation.app
• Custom domain for Spark dashboard: dash.filspark.com

Internal DevOps Dashboards

• Migrate remaining data sources from Protocol Labs Grafana
  • Filecoin Station Downloads API
  • Filecoin Station GitHub releases (not used)
  • Filecoin Station Plausible
• Meridian
• SPARK
• SPARK DB
• Station
• Migrate all alerts and ensure they send notifications to SpaceMeridian workspace
  • @juliangruber
    • Meridian
      • Available Balance
      • Balances (FIL) alert
      • Rounds alert
    • Station
      • All Stations Running alert
  • @bajtos - Spark
    • spark-api internal errors alert
    • spark-publish load alert
    • Unpublished measurements (Max Age) alert
    • SPARK DB capacity alert
    • spark-api request concurrency alert
    • spark-api response times alert
    • spark-stats internal errors alert
• Decide what to do with Station Public and how to merge those charts with the new Station Public Dashboard we are building

Cleanup

Remove Station/Spark/Voyager datasources from Protocol Labs Grafana account - see https://protocollabs.grafana.net/connections/datasources:

• Filecoin Station Downloads API
• Filecoin Station Fly.io
  
• Filecoin Station GitHub releases
• Filecoin Station InfluxDB
• Filecoin Station Plausible
• Filecoin Station Spark PostgreSQL
• Spark InfluxDB
• SPARK Stats

Close or rework older Grafana-related tasks:

• Build a Public Spark Dashboard in our new Grafana account #72
• Public Spark Dashboard MVP #57

[juliangruber]

Station dashboard was migrated https://spacemeridian.grafana.net/d/F2UmmYnVk/station?orgId=1

[juliangruber]

Next is to migrate the alerts.

@bajtos let's discuss this in the kickoff call, or do you have an idea how we could best do this?

[bajtos]

Next is to migrate the alerts.

@bajtos let's discuss this in the kickoff call

SGTM 👍🏻

or do you have an idea how we could best do this?

Not from the top of my head 😞

[juliangruber]

@bajtos should we go ahead and remove the datasources from the PL instances? I don't see why not, now that we have ported all dashboards

[bajtos]

IMO, we need to migrate the alerts first, as these alerts are reading data from the datasources, right?

[juliangruber]

great point!

[juliangruber]

Then I'll unsassign myself for the time being :)

[bajtos]

I migrated Spark alerts to SpaceMeridan's Grafana instance.

I am assigning the issue back to you, @juliangruber, so that you can perform the final cleanup. I don't have admin permissions to PL Grafana. Therefore, I cannot perform actions like deleting data sources.

In addition to "Remove Station/Spark/Voyager datasources", I think we should also remove our dashboards and alerts. WDYT?

[juliangruber]

All completed