Upload to grafana.com

[seriyps]

Hi,

Can those dashboards be uploaded to https://grafana.com/dashboards ?

[deadtrickster]

Yes but some of them have to be updated first :-(.
Uploaded allocators https://grafana.com/dashboards/6104 and Elli https://grafana.com/dashboards/6107.

[deadtrickster]

Grafana.com says this: "Dashboards from Grafana v3.0 or earlier cannot be uploaded."

If i understand it right upgrade is just import/export to/from recent enough version

[seriyps]

I have an instance of Grafana v5, so, I can try that

[seriyps]

But in my case it doesn't show some panels:

Uptime uses

process_uptime_seconds{instance="$node"}

And OS memory uses

process_virtual_memory_bytes{instance="$node"}
process_resident_memory_bytes{instance="$node"}

And I can't find them in the output from my node (prometheus 4.1.0).

Sorry for reporting this not in a separate ticket. Tell me if I need to create another one

[deadtrickster]

I guess we can reuse existing screenshots.

Perhaps you don't have prometheus_process_collector installed?

[seriyps]

Perhaps you don't have prometheus_process_collector installed?

Yep, that's true. Didn't knew about it. May we use erlang_vm_statistics_wallclock_time_milliseconds for uptime?

[seriyps]

Answer to myself:

Following config may show the uptime using erlang_vm_statistics_wallclock_time_milliseconds:

{
  "cacheTimeout": null,
  "colorBackground": false,
  "colorValue": false,
  "colors": [
    "rgba(245, 54, 54, 0.9)",
    "rgba(237, 129, 40, 0.89)",
    "rgba(50, 172, 45, 0.97)"
  ],
  "datasource": "XXX",
  "decimals": 1,
  "format": "dtdurationms",
  "gauge": {
    "maxValue": 100,
    "minValue": 0,
    "show": false,
    "thresholdLabels": false,
    "thresholdMarkers": true
  },
  "gridPos": {
    "h": 7,
    "w": 6,
    "x": 18,
    "y": 0
  },
  "id": 16,
  "interval": null,
  "links": [],
  "mappingType": 1,
  "mappingTypes": [
    {
      "name": "value to text",
      "value": 1
    },
    {
      "name": "range to text",
      "value": 2
    }
  ],
  "maxDataPoints": 100,
  "nullPointMode": "connected",
  "nullText": null,
  "postfix": "",
  "postfixFontSize": "50%",
  "prefix": "",
  "prefixFontSize": "50%",
  "rangeMaps": [
    {
      "from": "null",
      "text": "N/A",
      "to": "null"
    }
  ],
  "sparkline": {
    "fillColor": "rgba(31, 118, 189, 0.18)",
    "full": false,
    "lineColor": "rgb(31, 120, 193)",
    "show": false
  },
  "tableColumn": "",
  "targets": [
    {
      "expr": "erlang_vm_statistics_wallclock_time_milliseconds{instance=\"$node\"}",
      "intervalFactor": 2,
      "legendFormat": "",
      "refId": "A",
      "step": 4,
      "format": "time_series"
    }
  ],
  "thresholds": "",
  "title": "Uptime",
  "type": "singlestat",
  "valueFontSize": "80%",
  "valueMaps": [
    {
      "op": "=",
      "text": "N/A",
      "value": "null"
    }
  ],
  "valueName": "current"
}

highlights: "expr": "erlang_vm_statistics_wallclock_time_milliseconds{instance=\"$node\"}" Unit set to "duration (ms)" and Stat to "Current"

[deadtrickster]

maybe be add it as BEAM dashboard variant for people that don't use prometheus_process_collector? btw why you don't use it?

[seriyps]

@deadtrickster just don't want to add one more native dependency for just 2 metrics that I'm interested in