Bailongma

TDengine adatper for Prometheus, and others will happen soon.

For simple use case, just run:

sh bailongma

For all options, type --help for details.

Long usage here:

```sh TDengine adapter for prometheus

USAGE: bailongma [OPTIONS]

FLAGS: --help Prints help information

-V, --version
        Prints version information

OPTIONS: -c, --chunk-size Sql chunk size.

        The larger your table column size is, the small chunk should be setted. [default: 600]

-h, --host <host>
        TDengine host IP or hostname [default: localhost]

-l, --level <level>
        Debug level [default: info]

-L, --listen <listen>
        Listen to an specific ip and port [default: 0.0.0.0:10203]

-C, --max-connections <max-connections>
        Max TDengine connections

        - in concurrent cases, use max as 50000 - for common use, set it as 5000 [default:
        50000]

-M, --max-memory <max-memory>
        Max memroy, unit: GB [default: 50]

-P, --password <password>
        TDengine password [default: taosdata]

-p, --port <port>
        TDengine server port [default: 6030]

-u, --user <user>
        TDengine user [default: root]

-w, --workers <workers>
        Thread works for web request [default: 10]

```

Example:

sh bailongma -h tdengine -p 6030 -u root -P taospass --listen 0.0.0.0:10101

It will listen to port 10101(default is 10230).

Configure in prometheus.yml:

yaml remote_write: - url: "localhost:10101/adapters/prometheus/write" remote_read: - url: "localhost:10101/adapters/prometheus/read"

The default database is prometheus, use query option database to modify this, configuration file is like:

yaml remote_write: - url: "localhost:10101/adapters/prometheus/write?database=prom1" remote_read: - url: "localhost:10101/adapters/prometheus/read?database=prom1"

Build and Install

sh cargo build cargo install --path .

blm-bench-prom

cargo build will also produce a benchmark tool.

sh blm-bench-prom http://127.0.0.1:10230/adapters/prometheus/test5 \ --points 10 \ --metrics 10 \ --interval 1000 \ --chunks 1000 \ --samples 100