mongodb-grafana-plugin

this repo tries to modernize the good work from the mongodb-grafana repo.

the plugin is a grafana data source plugin. it is a full stack app, it has to, because a webapp can't load a mongodb driver binary in a browser. there are two parts in the plugin,

• a frontend
  • used by the grafana UI
  • help configure a database connection and collect user queries
• a backend
  • run outside of grafana as a standalone node.js app
  • receive the queries from the frontend
  • call a mongodb driver to execute the queries
  • return as either tabular or timeseries data frames

installation

see installation steps here

query

th plugin supports all mongodb aggregation pipeline operators

query types

there are two types of query the plugin supports, table or timeserie.

query schema

after the frontend UI collects a query, it transforms and sends data of the following schema to its backend. an example query payload to the backend can be found at code

  {
    collection: "collection name",
    aggregations: [
      {
        pipeline stage (s)
      }
    ],
  }

a query is a javascript object, NOT a json object.

the Collection Name field, (mapped to collection in the above payload), specifies the name of the collection to perform an aggregation.

the Aggregation Pipeline field, (mapped to aggregations), can have multiple aggregation steps. if you're familiar with MongoDB Compass, you can contruct your aggregations in the tool, then copy/paste the aggregration's Node.js export to the field - it's just simple as that. the supported data types are listed below.

  string: "string",
  date: new Date(),
  id: new ObjectId("573a1393f29313caabcdc50e"),
  bool: true,
  number: 12345.4,
  expandable1: $from,
  expandable2: $to,
  expandable3: $intervalMs

timeserie in detail

• $from, $to, $intervalMs are grafana expandable variables for using as a time boundary and an interval
• __name, __value, __timestamp are the plugin's internal variables to construct a named grafana timeserie data frame
• in this example and in its aggregations array, all 6 variables mentioned above are used. it $group by month and $sum how many movies were released during a month of a year.

sample dashboard

an example dashboard can be found at here

visualizations above courtesy from mongodb atlas's default sample_mflix database which collects movies from early 20th century all the way to 2016

contribution

you're welcomed to contribute to the repo in any way.