influxdb-cpp-2.0

A header-only C++ query & write client for InfluxDB 2.0.

• Support versions:
  • InfluxDB v2.0
  • Incompatible with versions 1.x
  • Check yourself while using future versions.

Why use influxdb-cpp?

• Exactly-small:
  • Less than 300 lines and only 10KB+.
• Easy-to-use:
  • It's designed to be used without extra studies.
• Easy-to-assemble:
  • Only a tiny header file needs to be included.
• No-dependencies:
  • Unless STL and std C libraries.

Examples

Before using

• The very simple thing you should do before using is only:

  #include "influxdb.hpp"

Writing example

• You should refer to the write syntax while writing series(metrics).

  measurement[,tag-key=tag-value...] field-key=field-value[,field2-key=field2-value...] [unix-nano-timestamp]
  

• You can rapidly start writing series according to the following example.

  influxdb_cpp::server_info si("127.0.0.1", 8086, "org", "token", "bucket");
  influxdb_cpp::builder()
      .meas("foo")
      .tag("k", "v")
      .tag("x", "y")
      .field("x", 10)
      .field("y", 10.3, 4)
      .field("z", 10.3456)
      .field("b", !!10)
      .timestamp(1512722735522840439)
      .post_http(si);

  • Remarks:
    • 3rd parameter precision of field() is optional for floating point value, and default precision is defined in the top of the header.
    • The token must have access to write (and/or read) from the specified organization and bucket.
    • Although bucket is an optional parameter for a server info object, it is required for database writes.

• The series sent is:

  foo,k=v,x=y x=10i,y=10.30,z=10.346,b=t 1512722735522840439
  

• Bulk/batch/multiple insert also supports:

  influxdb_cpp::builder()
      .meas("foo")  // series 1
      .field("x", 10)
  
      .meas("bar")  // series 2
      .field("y", 10.3)
      .post_http(si);

• The series sent are:

  foo x=10i
  bar y=10.30
  

Query example

• And you can query series using the Flux query language.

  influxdb_cpp::server_info si("localhost", 8086, "spartans", "my_token", "");
  
  string resp;
  string query("from(bucket: \\\"SR-20\\\")|> range(start: -5m)|>filter(fn: (r)=>r[\\\"_measurement\\\"] == \\\"cpu\\\")|>filter(fn: (r) => r[\\\"_field\\\"] == \\\"usage_user\\\")");
  
  influxdb_cpp::query(resp, query, si);

• NOTE: The query is sent in a JSON key-pair, therefore there must be no newlines/tabs, and " must be double-escaped as \\\"!

• The results are returned in CSV format. A column, table, separates different measurements

Remarks for Windows

• You should init socket environment by yourself under Windows.
  • FYR: MSDN doc for WSAStartup

TODO

• Add more test cases.
• Supports DSN initializatin for server_info.
• Do not need to connect every time.

Misc

• Please feel free to use influxdb-cpp.
• Looking forward to your suggestions.
• If your project is using influxdb-cpp, you can show your project or company here by creating a issue or let me know.