This adds rules for validating the dashboards using the official upstream schema.

I've marked this as a draft because, after having seen how the flow of this tool actually works, it seems like a sloppy way to integrate this kind of logic - the possibility of false positives, and the messiness of errors, make it maaaaaaybe a good addition (it's good enough that we no longer allow Grafana dashboards to enter grafana/grafana/devenv without passing validation), but still not completely canonical and reliable. OTOH, that isn't worse than e.g. false positives arising from expecting only Prom queries.

Either way, i figured it was worth putting the PR up at minimum to show how this kind of addition could be made. My guess is that a larger refactor of the application to treat validation as a prerequisite to linting would be a better architecture, and would also probably set us up for when we have Go structs that represent dashboard structures.

As mentioned in the $title, this PR enables templated variable substitution in panel promql expressions.

Example: https://github.com/kubernetes-monitoring/kubernetes-mixin/blob/master/dashboards/network-usage/cluster-total.libsonnet#L395

Signed-off-by: Arunprasad Rajkumar [email protected]

Hello, I my company, we have a rule that the first panel of a dashboard should be named "Dashboard Info" and should contain some panels with PoC, ... I was hoping, thanks to this PR to replace my current jq based solution with a Custom Rule integrated to the linter. My intent is to gather feedback on the idea. Before merging I would like to add some test cases, to at least validate the case when a rule implements several interfaces (dashboard, panel, target linter). Thanks again in advance for your inputs Gabriel

Another solution to the problem shared in https://github.com/grafana/dashboard-linter/pull/18 Tested successfully on the HEAD of https://github.com/kubernetes-monitoring/kubernetes-mixin.git (removing the sed line in the Makefile) @tomwilkie @arajkumar

Reference issue: https://github.com/grafana/dashboard-linter/issues/52

This rule checks if each dashboard has:

• A default time interval of 1h
• A default refresh interval of 5m

Add a linter to golangci-lint (first build will check to validate it actually works). Then I will format the code to make it pass.

Btw, in another project I use this setup: linters: enable: - bodyclose - deadcode - errcheck - goconst - gocritic - goerr113 - gofmt - goprintffuncname - gosec - gosimple - govet - ineffassign - misspell - nakedret - nolintlint - prealloc - exportloopref - staticcheck - structcheck - testpackage - unconvert - unused - varcheck - whitespace - gocognit - godox - ifshort - importas - forbidigo

linters-settings: gocognit: min-complexity: 15 forbidigo: forbid: - ^print.$ - 'fmt.Print.'

@tomwilkie , do you want me to explore activating those too?

Hello, This should cover a case I have internally, with a query like sum(increase(foo{...}[$__range])) interval, which I believe should not be rejected by the linter. Let me know if you have concerns. Gabriel

Reference issue: https://github.com/grafana/dashboard-linter/issues/51

This rule checks that each panel must have a valid unit assigned. The unit is extracted from dashboard json in the following manner:

• Parse overrides to check if any override unit is present
• If no overrride unit - then extract standard option unit
• Check if unit is defined and is valid

Valid units are extracted from here: https://github.com/grafana/grafana/blob/main/packages/grafana-data/src/valueFormats/categories.ts

Considerations:

• For a panel with multiple fields and panels - the last override unit will be picked

Sample output:

Bumps github.com/prometheus/prometheus from 0.36.0 to 0.36.1.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Bumps github.com/stretchr/testify from 1.7.1 to 1.7.2.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Summary of the issue

When you have a multiple choice variable in your dashboard, linter will throw an error.

Cause of the issue

TemplateValue struct is expecting 2 fields as string (https://github.com/grafana/dashboard-linter/blob/main/lint/lint.go#L35), but multiple choice variables are arrays:

  "templating": {
    "list": [
      {
        "current": {
          "selected": true,
          "text": [
            "All"
          ],
          "value": [
            "$__all"
          ]
        },
...

Expected behaviour

Linter shouldn't throw an error when the dashboard has multiple choice variables.

How to replicate

Just add a multiple choice variable to your dashboard. For example

{
  "annotations": {
    "list": [
      {
        "builtIn": 1,
        "datasource": "-- Grafana --",
        "enable": true,
        "hide": true,
        "iconColor": "rgba(0, 211, 255, 1)",
        "name": "Annotations & Alerts",
        "target": {
          "limit": 100,
          "matchAny": false,
          "tags": [],
          "type": "dashboard"
        },
        "type": "dashboard"
      }
    ]
  },
  "editable": true,
  "fiscalYearStartMonth": 0,
  "graphTooltip": 0,
  "iteration": 1649669111850,
  "links": [],
  "liveNow": false,
  "panels": [
    {
      "fieldConfig": {
        "defaults": {
          "color": {
            "mode": "palette-classic"
          },
          "custom": {
            "axisLabel": "",
            "axisPlacement": "auto",
            "barAlignment": 0,
            "drawStyle": "line",
            "fillOpacity": 0,
            "gradientMode": "none",
            "hideFrom": {
              "legend": false,
              "tooltip": false,
              "viz": false
            },
            "lineInterpolation": "linear",
            "lineWidth": 1,
            "pointSize": 5,
            "scaleDistribution": {
              "type": "linear"
            },
            "showPoints": "auto",
            "spanNulls": false,
            "stacking": {
              "group": "A",
              "mode": "none"
            },
            "thresholdsStyle": {
              "mode": "off"
            }
          },
          "mappings": [],
          "thresholds": {
            "mode": "absolute",
            "steps": [
              {
                "color": "green",
                "value": null
              },
              {
                "color": "red",
                "value": 80
              }
            ]
          }
        },
        "overrides": []
      },
      "gridPos": {
        "h": 9,
        "w": 12,
        "x": 0,
        "y": 0
      },
      "id": 2,
      "options": {
        "legend": {
          "calcs": [],
          "displayMode": "list",
          "placement": "bottom"
        },
        "tooltip": {
          "mode": "single",
          "sort": "none"
        }
      },
      "title": "Panel Title",
      "type": "timeseries"
    }
  ],
  "refresh": "",
  "schemaVersion": 35,
  "style": "dark",
  "tags": [],
  "templating": {
    "list": [
      {
        "current": {
          "selected": true,
          "text": [
            "All"
          ],
          "value": [
            "$__all"
          ]
        },
        "hide": 0,
        "includeAll": true,
        "label": "Variable",
        "multi": true,
        "name": "query0",
        "options": [
          {
            "selected": true,
            "text": "All",
            "value": "$__all"
          },
          {
            "selected": false,
            "text": "foo",
            "value": "foo"
          },
          {
            "selected": false,
            "text": "bar",
            "value": "bar"
          }
        ],
        "query": "foo,bar",
        "queryValue": "",
        "skipUrlSync": false,
        "type": "custom"
      }
    ]
  },
  "time": {
    "from": "now-6h",
    "to": "now"
  },
  "timepicker": {},
  "timezone": "",
  "title": "New dashboard",
  "version": 0,
  "weekStart": ""
}

Architecture. telegraf+influxdb+grafana Requirements. Is there a statistics panel to display server information, e.g., 20 servers in total, 4 abnormal and the rest normal.

Dashboard variable like below is not flagged by linter. You can see that UID is assigned specific value pjNv0fYGz. Such dashboard is not portable across different grafana environments. It would be nice for a linter to flag such case.

      {¬
        "current": {¬
          "selected": false,¬
          "text": "X",¬
          "value": "X"¬
        },¬
        "datasource": {¬
          "type": "prometheus",¬
          "uid": "pjNv0fYGz"¬
        },¬
        "definition": "definition",¬
        "hide": 0,¬
        "includeAll": false,¬
        "label": "Customer ID",¬
        "multi": false,¬
        "name": "customerid",¬
        "options": [],¬
        "query": {¬
          "query": "PromQL",¬
          "refId": "StandardVariableQuery"¬
        },¬
        "refresh": 2,¬
        "regex": "",¬
        "skipUrlSync": false,¬
        "sort": 0,¬
        "type": "query"¬
      }

I have a constant variable in a dashboard and it looks like it gets flagged by linter

     {
        "hide": 2,¬
        "name": "instance",¬
        "query": ".*",¬
        "skipUrlSync": false,¬
        "type": "constant"¬
      }

$HOME/go/bin/dashboard-linter lint mydashboard.json

[❌] Dashboard 'My dashboard' instance template should use datasource '$datasource', is currently ''

Should it be?

Resolves #56.

Fairly simple, and literal check that finds any metric/series containing _bucket and ensures that it is inside of a function call to histogram_quantile.

There was some internal slack discussion about forthcoming features which may need to be accounted for in the future, but we will wait until those are finalized before addressing them in the rules.

All the queries that use a template variable which allows multiselection should be aggregated:

• Either by using an aggregator function in PromQL or LogQL
• Or by using an aggregation option from the Grafana panel options

This will avoid a situation where we see multiple repeating values unintentionally in a panel like "Single stat".

This rule should also be limited to stat or panels that are used to show aggregated stats.