Create a Business Dashboard in Grafana for release management

We firmly believe that Grafana is not just about observability; Grafana can help build fully functional, user-ready web applications with the right tools, like the Business Suite plugins.

This article aims to be one example. Release dashboard is an internal Volkov Labs project available for the public to view. We use it daily to monitor our plugin release schedule and community interest by using various metrics such as number of downloads, open issues, etc.

Below is the screenshot of the main dashboard with noted visualization panels.

Release dashboard sectioned.

You can jump into the specific panel configuration using the listing below:

• Lifecycle panel
• Open issues panel
• Toolkit panel
• Weekly downloads panel
• Total downloads panel
• Total downloads weekly panel
• Open issues over time panel
• Release Calendar panel

Data Flow

Let's delve into the details of our data flow. This brief explanation will equip you with a deeper understanding of where our data comes from and how it's collected, making you more informed and knowledgeable about our processes.

Data flow schema.

The data that we use in our Release dashboard is collected from three different sources:

• From grafana.com using Grafana API and hourly CRON script.
• From github.com using GitHub API and hourly CRON script.
• Manual entry. When the release is published, release manager manually enters all required information via the Grafana dashboard using the Business Forms panel.

The Grafana dashboard to enter release details into the data source.

Timescale database

We use Timescale database to store all collected data. The primary reason for our choice in the Release project is Timescale's ability to combine table structured data with time-series format and allow querying both simultaneously.

info

For more reasons to choose Timescale, please review this blog post.

As you can see on the data flow schema above, we primarely use three tables: grafana, github, and releases.

Grafana plugins

Below is a part of the grafana table to illustrate the collected data about plugins and number of downloads from Grafana catalog.

Collected data about plugins and number of downloads.

• id. A unique Grafana plugin identifier in the integer format.
• slug. A unique Grafana plugin identifier in the text format.
• time. The time when a data is collected.
• downloads. Number of downloads.
• diff. The number of downloads increase since the last time the data was collected.

GitHub issues

Below is a part of the github table to illustrate the collected data about issue, stars, and forks from our GitHub organization.

Collected data about issue, stars, and forks from our GitHub organization.

• time. The time when a GitHub project data is collected.
• id. A unique GitHub project identifier in the integer format.
• name. A unique GitHub project identifier in the text format.
• forks. A number of project's forks.
• stars. A number of project's stars.
• issues. A number of project's open issues.

Business Suite releases

Below is a part of the releases table to illustrate the collected data about Business Suite releases.

Collected data about Business Suite releases.

• id. A unique row identifier.
• start_date. The data when a release is first available.
• end_date. The start_date + 100 days.
• title. The plugin name + version.
• labels. The Grafana version compatible with the release.
• color. A unique plugin identifier.
• description. The most significant release features are in HTML format.

Any panel configuration

Any Grafana dashboard consists of various visualization panels. Though the variety of panel types is staggering, the configuration of any kind can be summarized in the following steps.

Step	Notes
1. Data source selection	One or multiple data sources per one panel. You can reference the other panel on the same dashboard as a data source to avoid unnecessary duplicative query runs and, therefore, improve performance
2. Specify query	One or multiple queries per one panel. You can use the query Builder or Code mode to specify a query
3. Apply transformations	One or multiple transformations per panel
4. Data review	Use the Table View control at the top to switch between the data view and the panel view. You can enable a transformation using an eye icon to change from the original data frame to transformed and back. It is an optional step used for troubleshooting
5. Select a visualization panel from a list	Choose from the core visualization panels or pre-installed community panels
6. Configure panel properties	Every panel has common and specific properties

Any visualization panel configuration steps.

Grafana dashboard

Lifecycle panel

The panel shows our products: Business Alerting, open-source Business Suite, and the Environment data source.

For each, you can see the number of days since the last release and how long is until the next planned release. We aim to have a new release every 100 days for each plugin and so far have been successful with that goal.

Lifecycle panel using the Business Charts panel (Bar visualization).

Below is the Lifecycle panel configuration.

Step	Lifecycle panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs two SQL queries: Query A returns the product name(title) and the number of days since the last release has been published. Query B returns the product name(title) and the number of days until the next planned release.
3. Apply transformations	We use the Join by field transformation to merge two data frames (results of the Query A and Query B) into one data frame
4. Extracted data review
5. Select a visualization panel from a list	Business Charts panel
6. Configure panel properties	We use the Code Editor mode. See the code below

Lifecycle panel configuration.

• Query A:

WITH my_time as (SELECT color, max(start_date) as max_time FROM releases GROUP BY color)
SELECT a.title, now()::date-a.start_date::date as release
FROM releases a, my_time mt
WHERE a.color = mt.color and a.start_date = mt.max_time
order by title desc;

• Query B:

WITH my_time as (SELECT color, max(end_date) as max_time FROM releases GROUP BY color)
SELECT a.title, a.end_date::date-now()::date as milestone
FROM releases a, my_time mt
WHERE a.color = mt.color and a.end_date = mt.max_time
order by title asc

info

There are plently of ways to learn how to configure the Business Charts panel:

• Sandbox echarts.volkovlabs.io.
• Instruction on how to use the Sandbox.
• Visual Editor mode
• Blog posts
• Our video tutorials

• The Function parameter JavaScript code:

let titles = [];
let releases = [];
let milestones = [];

context.panel.data.series.map((s) => {
  titles = s.fields.find((f) => f.name === "title").values;
  releases = s.fields.find((f) => f.name === "release").values;
  milestones = s.fields.find((f) => f.name === "milestone").values;
});

return {
  tooltip: {
    trigger: "axis",
    axisPointer: {
      type: "shadow",
    },
  },
  legend: {},
  grid: {
    left: "3%",
    right: "4%",
    bottom: "3%",
    containLabel: true,
  },
  xAxis: {
    type: "value",
  },
  yAxis: {
    type: "category",
    data: titles,
  },
  series: [
    {
      name: "Since Release",
      type: "bar",
      stack: "total",
      label: {
        show: true,
      },
      emphasis: {
        focus: "series",
      },
      data: releases,
    },
    {
      name: "Until Milestone",
      type: "bar",
      stack: "total",
      label: {
        show: true,
      },
      emphasis: {
        focus: "series",
      },
      data: milestones,
    },
  ],
};

Open Issues panel

This visualization shows the number of currently open issues per each plugin. This metric highlights the plugins with the most interest from the Grafana community. At this moment, our Business Table panel is a star with 34 open issues, while the second place takes the Business Forms panel with only 8.

Open issues panel using the Business Charts panel (Pie visualization).

Below is the Open Issues panel configuration.

Step	Open Issues panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query which returns the number of open issues per plugin
3. Apply transformations	We use the Filter data by values transformation to keep plugins only open issues greater than 0. And then Sort by to order plugins by that number
4. Extracted data review
5. Select a visualization panel from a list	Business Charts panel
6. Configure panel properties	We use the Code Editor mode. See the code below

Open issues panel configuration.

• Query A:

SELECT name, title, last(issues, time)
FROM github join products on name=slug
GROUP BY name, title;

• The Function parameter JavaScript code:

const pieData = context.panel.data.series.map((s) => {
  const models = s.fields.find((f) => f.name === "title").values;
  const values = s.fields.find((f) => f.name === "last").values;

  return values.map((d, i) => {
    return { name: models[i], value: d };
  });
})[0];

const sum = pieData.reduce((prev, current) => prev + current.value, 0);

return {
  tooltip: {
    trigger: "item",
  },
  title: {
    text: `${sum}`,
    left: "center",
    top: "center",
    textStyle: {
      fontSize: context.panel.chart.getHeight() / 10,
    },
  },
  series: [
    {
      type: "pie",
      radius: ["30%", "70%"],
      roseType: "radius",
      avoidLabelOverlap: false,
      itemStyle: {
        borderRadius: 10,
        borderColor: "#fff",
        borderWidth: 2,
      },
      label: {
        show: true,
      },
      emphasis: {
        label: {
          show: true,
          fontSize: "16",
          fontWeight: "bold",
        },
      },
      labelLine: {
        show: true,
      },
      data: pieData,
    },
  ],
};

Toolkit panel

This panel shows the number of our plugins that have migrated to the latest Grafana versions. All our products migrate promptly. At this moment, all of them are at least on Grafana 11.1

Grafana Toolkit panel using the Business Charts panel (Pie visualization).

Below is the Grafana Toolkit panel configuration.

Step	Grafana Toolkit panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query which returns Grafana version for each plugin
3. Apply transformations	We use the Group by transformation to calculate the number of supported plugins by each Grafana version
4. Extracted data review
5. Select a visualization panel from a list	Business Charts panel
6. Configure panel properties	We use the Code Editor mode. See the code below

Grafana Toolkit panel configuration.

• Query A:

WITH my_time as (SELECT color, max(start_date) as max_time FROM releases GROUP BY color)
SELECT a.labels, 1 as count FROM releases a, my_time mt
WHERE a.color = mt.color and a.start_date = mt.max_time
order by 1 desc

• The Function parameter JavaScript code:

const pieData = context.panel.data.series.map((s) => {
  const models = s.fields.find((f) => f.name === "labels").values;
  const values = s.fields.find((f) => f.name === "count (sum)").values;

  return values.map((d, i) => {
    return { name: models[i], value: d };
  });
})[0];

const sum = pieData.reduce((prev, current) => prev + current.value, 0);

return {
  tooltip: {
    trigger: "item",
  },
  title: {
    text: `${sum}`,
    left: "center",
    top: "center",
    textStyle: {
      fontSize: context.panel.chart.getHeight() / 10,
    },
  },
  series: [
    {
      type: "pie",
      radius: ["30%", "70%"],
      roseType: "radius",
      avoidLabelOverlap: false,
      itemStyle: {
        borderRadius: 10,
        borderColor: "#fff",
        borderWidth: 2,
      },
      label: {
        show: true,
      },
      emphasis: {
        label: {
          show: true,
          fontSize: "20",
          fontWeight: "bold",
        },
      },
      labelLine: {
        show: true,
      },
      data: pieData,
    },
  ],
};

Weekly downloads panel

This panel is the first of the three, showing the number of our plugin downloads. This visualization focuses on the weekly downloads to illustrate community interest from one week to another.

Weekly downloads panel using the Timeseries panel.

Below is the Weekly downloads panel configuration.

Step	Weekly downloads panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query in the Time series format, which returns a date and then a number of weekly downloads for each of the specified in the query products
3. Apply transformations	We use the Organize fields by name transformation to convert database specific product names into user-friendly names
4. Extracted data review
5. Select a visualization panel from a list	Time series
6. Configure panel properties	Use the panel parameter to configure the appearance

Weekly downloads panel configuration.

• Query A:

select date_trunc('week', time) as time, sum(diff), slug from grafana
where $__timeFilter("time") and
slug in ('volkovlabs-echarts-panel', 'volkovlabs-form-panel',
'volkovlabs-rss-datasource', 'volkovlabs-image-panel',
'marcusolsson-calendar-panel', 'marcusolsson-dynamictext-panel', 'marcusolsson-static-datasource',
'volkovlabs-grapi-datasource', 'volkovlabs-variable-panel', 'volkovlabs-table-panel')
group by 1, 3
order by 1 asc;

Total downloads panel

This panel is the second of the three, showing the number of our plugin downloads. This visualizaiton focuses on the total downloads from the begining of time until now.

Total downloads panel using the Stat panel.

Below is the Total downloads panel configuration.

Step	Total downloads panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query which returns a last counted number of downloads for each specified plugin
3. Apply transformations	No need for transformation in this panel
4. Extracted data review
5. Select a visualization panel from a list	Stat
6. Configure panel properties	We use the Total Grafana function, to sum up downloads across all plugins

Total downloads panel configuration.

• Query A:

select last(downloads, time), slug from grafana
where $__timeFilter("time") and
slug in ('volkovlabs-echarts-panel', 'volkovlabs-form-panel',
'volkovlabs-rss-datasource', 'volkovlabs-image-panel',
'marcusolsson-calendar-panel', 'marcusolsson-dynamictext-panel', 'marcusolsson-static-datasource',
'volkovlabs-grapi-datasource', 'volkovlabs-variable-panel', 'volkovlabs-table-panel')
group by slug
order by 1;

Total downloads weekly panel

This panel is the third of the three, showing the number of our plugin downloads. It provides details for plugin downloads every week. We can see steadily growing interest in all our Business Suite plugins.

Total downloads weekly panel using the Timeseries panel.

Below is the Total downloads weekly panel configuration.

Step	Total downloads weekly panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query in the Time series format, which returns a date and the number of plugin downloads weekly
3. Apply transformations	We use the Organize fields by name transformation to convert database specific product names into user-friendly names
4. Extracted data review
5. Select a visualization panel from a list	Time series
6. Configure panel properties	Use the panel parameter to configure the appearance

Total downloads weekly panel configuration.

• Query A:

select date_trunc('week', time) as time, last(downloads, time), slug from grafana
where $__timeFilter("time") and
slug in ('volkovlabs-echarts-panel', 'volkovlabs-form-panel',
'volkovlabs-rss-datasource', 'volkovlabs-image-panel',
'marcusolsson-calendar-panel', 'marcusolsson-dynamictext-panel', 'marcusolsson-static-datasource',
'volkovlabs-grapi-datasource', 'volkovlabs-variable-panel', 'volkovlabs-table-panel')
group by 1, 3
order by 1 asc;

Open issues over time panel

This panel accompanies the open issues snapshot, allowing for deeper community interest analysis.

Open issues over time panel using the Timeseries panel.

Below is the Open issues over time panel configuration.

Step	Open issues over time panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query in the Time series format, which returns a date and then an average number of open issues weekly
3. Apply transformations	We use the Organize fields by name transformation to convert database specific product names into user-friendly names
4. Extracted data review
5. Select a visualization panel from a list	Time series
6. Configure panel properties	Use the panel parameter to configure the appearance

Open issues over time panel configuration.

• Query A:

select date_trunc('week', time) as time, avg(issues), name from github
where $__timeFilter("time")
group by 1, 3
order by 1 asc;

Release Calendar panel

On this panel, releases have been published in the calendar format. When a mouse hovers over a plugin name, you can see the release details in HTML format.

Release Calendar panel using the Business Calendar panel.

Below is the Release Calendar panel configuration.

Step	Release Calendar panel configuration
1. Data source selection	Timescale
2. Specify query	The panel runs one SQL query, which returns all data elements necessary to display an informative calendar
3. Apply transformations	No transformation needed
4. Extracted data review
5. Select a visualization panel from a list	Business Calendar
6. Configure panel properties	Use the Data parameter category to map data frame columns with the Business Calendar data elements

Release Calendar panel configuration.

• Query A:

SELECT id, start_date AT TIME ZONE 'UTC' as start_date,
end_date AT TIME ZONE 'UTC' as end_date, title,
description, labels, color
FROM releases;

To sum up

This article explained how we created the Release dashboard, an excellent example of using Grafana for Business purposes. The beautiful and story-telling visualizations are complemented with a data entry form (possible with the Business Form panel), the tool allowing direct user interaction with the data source.

info

You can learn more about the Business Forms panel in the tutorials section of our documentation.

Always happy to hear from you

• Subscribe to our YouTube Channel and leave your comments.
• Become a Business Suite sponsor.

  Enroll in Business Suite Enterprise