Simple HTTP API backend server for Grafana based on Deno. Supported by Grafana Simple JSON Datasource. This will be useful for prototyping dashboards, demo dashboards and while building plugins.
- How to use
- Features
- Custom TimeSeries patterns
- Custom Table Data
- CSV Datasource
- Local Installation
- Notes
- License
The backend server is hosted in the following url
Above backend needs to be connected with Grafana Simple JSON Datasource. Example shown here. Once datasource configured, you can start writing queries directly.
You can also fork the repo and connect with vercel to get your own instance of the URL.
Note: Instead the hosted version, you can also use your own instance for production purposes
- Custom timeseries patterns
- CSV Datasource
- Ad-hoc table data
- LED text
For more details about the use cases, refer the screenshots here.
| Series Name | Result |
|---|---|
| foo | Single random walk series called foo |
| bar | Single random walk series called bar |
| RandomWalk() | Single random walk series with random name |
| RandomWalk(10) | 10 random walk series with random names |
| RandomWalk(3,Server) | 3 random walk series with the name 'Server 1', 'Server 2' and 'Server 3' |
| FlatLine() | Flat series with constant 0 as value over the period with random series name |
| FlatLine(16) | Flat series with constant 16 as value over the period with random series name |
| FlatLine(7,Rainbow) | Flat series with constant 7 as value over the period with 'Rainbow' as series name |
| Step() | Random series starting from 0 and increment by 1 for each data point |
| Step(10) | Random series starting from 10 and increment by 1 for each data point |
| Step(10,Speed) | Series called 'Speed' starting from 10 and increment by 1 for each data point |
| Step(10,Speed,5) | Series called 'Speed' starting from 10 and increment by 5 for each data point |
| Step(10,,5) | Random series starting from 10 and increment by 5 for each data point |
| Pattern(HeartBeat,0,0,0,-3,-2,4,0,0) | Repeat the pattern 0,0,0,-3,-2,4,0,0 as series 'Heartbeat' over the time range |
| Pattern(,0,1,2,1) | Repeat the pattern 0,1,2,1 as random series over the time range |
| Patterns([Foo,4,3,2,1,2,3][Bar,1,2,3,2]) | Repeat both the patterns over the time range |
| Maths(Foo,1 (2*$i)) | Return series called Foo using the math series 1 (2*$i) where $i is the index. |
| Maths(Bar,sin($i deg) * pow($i,3)) | Return series called Bar using the math series sin($i deg) * pow($i,3) where $i is the index. |
| Expression(Foo,multiply:0.1,cos,abs,max:0.3) | Series foo which is equivalent of max(abs(cos(x*0.1)),0.3) where x is the index of the datapoint starting from 0. Multiple expressions can be added. Valid expressions are random, add, minus, multiply, divide, abs, pow, sqrt, max, min, sin, cos, tan, ceil, floor and round. |
| Matrix(Foo,10,20,[6,7,1,2,3,4,5][1,2,0,4,5]) | Create a series call Foo after the fist 10 timestamps. shifted by 20 points down. Rest of the series within square brackets are datapoints for each timestamp. Given example will have 2 timestamp points. |
| LED(GRAFANA,10) | Create series of datapoints representing word GRAFANA in LED format.Spaces and alphabets are accepted. If you need space before/after the word prefix/suffix with *. Second parameter is y offset. In the example entire text is shifted by 10 points in y axis |
| Function | Sample Query | Description |
|---|---|---|
| start-after | Expression(Foo,sin,start-after:2) | Series foo will start after 2 null values |
| stop-after | Expression(Foo,sin,stop-after:2) | Series foo will stop after values and produce null for remaining series |
For custom table data, follow the below syntax
Custom(columnCount,...columns,...rows)
Examples given below
| Format | Description |
|---|---|
Custom(1,countries,india,china,england,united kingdom) |
1 column table with column name as countries. Contains india, china, england, united kingdom as rows |
Custom(2,countries,city,india,chennai,england,london) |
2 columns table with Countries and City as column names. Remaining items will spread as row items respectively. |
Custom(3,countries,rank:number,capital,india,1,delhi,england,3,london) |
3 column table where second column in number format |
You can call any csv url and the api server will return the data to grafana.
| Format | Sample Query | Description |
|---|---|---|
CSV(YOUR_CSV_URL) |
CSV(https://gist.githubusercontent.com/yesoreyeram/64a46b02f0bf87cb527d6270dd84ea47/raw/f4d021529650be41f863cda73e41a6242fb2ea27/sample.csv) | Read the sample.csv and returns to grafana |
CSV(YOUR_CSV_URL, headers:column1:column2) |
CSV(https://gist.githubusercontent.com/yesoreyeram/64a46b02f0bf87cb527d6270dd84ea47/raw/85bfe906d5eda20611496854e22ecc57a901bf4c/sample-without-header.csv, headers:Country Name:city Name) | If the csv file doesn't contain headers, provide the headers as option. You can specify n number of columns with : separated. |
More examples can be found here
- Deno 1.x or higher
There are couple of ways you can start the server
- Running directly from url
deno run --allow-net https://raw.githubusercontent.com/yesoreyeram/grafana-utils/master/mod.ts --port 8080
- Running from local
git clone https://github.com/yesoreyeram/grafana-utils.git
cd grafana-utils
deno run --allow-net mod.ts --port 8080
There are few command line flags you can pass while starting the server to customize the server.
| Param | Description |
|---|---|
--allow-net |
This is security flag required by deno. Otherwise, server will not able to access the port. |
--port |
HTTP port to start the server. If not provided, server will try to start at port 8080 |
Timeseries data are set for auto granularity to stop abuse of the platform.
- More than 13 months range -> 1 week granularity;
- Less than 13 months range -> 1 day granularity;
- Less than 40 days range -> 1 hour granularity;
- Less than 2 days range -> 1 minute granularity;
Grafana utils is licensed under the Apache 2.0 License.