Skip to content

pabigot/influxdb-client-go

 
 

Repository files navigation

InfluxDB Client Go

CircleCI codecov License Slack Status

This repository contains the reference Go client for InfluxDB 2.

Note: Use this client library with InfluxDB 2.x and InfluxDB 1.8+ (see details). For connecting to InfluxDB 1.7 or earlier instances, use the influxdb1-go client library.

Features

Documentation

This section contains links to the client library documentation.

Examples

Examples for basic writing and querying data are shown below in this document

There are also other examples in the API docs:

How To Use

Installation

Go 1.13 or later is required.

  1. Add the client package your to your project dependencies (go.mod).
    go get github.com/influxdata/influxdb-client-go/v2
  2. Add import github.com/influxdata/influxdb-client-go/v2 to your source code.

Basic Example

The following example demonstrates how to write data to InfluxDB 2 and read them back using the Flux language:

package main

import (
    "context"
    "fmt"
    "time"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Use blocking write client for writes to desired bucket
    writeAPI := client.WriteAPIBlocking("my-org", "my-bucket")
    // Create point using full params constructor 
    p := influxdb2.NewPoint("stat",
        map[string]string{"unit": "temperature"},
        map[string]interface{}{"avg": 24.5, "max": 45.0},
        time.Now())
    // write point immediately 
    writeAPI.WritePoint(context.Background(), p)
    // Create point using fluent style
    p = influxdb2.NewPointWithMeasurement("stat").
        AddTag("unit", "temperature").
        AddField("avg", 23.2).
        AddField("max", 45.0).
        SetTime(time.Now())
    writeAPI.WritePoint(context.Background(), p)
    
    // Or write directly line protocol
    line := fmt.Sprintf("stat,unit=temperature avg=%f,max=%f", 23.5, 45.0)
    writeAPI.WriteRecord(context.Background(), line)

    // Get query client
    queryAPI := client.QueryAPI("my-org")
    // Get parser flux query result
    result, err := queryAPI.Query(context.Background(), `from(bucket:"my-bucket")|> range(start: -1h) |> filter(fn: (r) => r._measurement == "stat")`)
    if err == nil {
        // Use Next() to iterate over query result lines
        for result.Next() {
            // Observe when there is new grouping key producing new table
            if result.TableChanged() {
                fmt.Printf("table: %s\n", result.TableMetadata().String())
            }
            // read result
            fmt.Printf("row: %s\n", result.Record().String())
        }
        if result.Err() != nil {
            fmt.Printf("Query error: %s\n", result.Err().Error())
        }
    }
    // Ensures background processes finishes
    client.Close()
}

Options

The InfluxDBClient uses set of options to configure behavior. These are available in the Options object Creating a client instance using

client := influxdb2.NewClient("http://localhost:8086", "my-token")

will use the default options.

To set different configuration values, e.g. to set gzip compression and trust all server certificates, get default options and change what is needed:

client := influxdb2.NewClientWithOptions("http://localhost:8086", "my-token", 
    influxdb2.DefaultOptions().
        SetUseGZip(true).
        SetTLSConfig(&tls.Config{
            InsecureSkipVerify: true,
        }))

Writes

Client offers two ways of writing, non-blocking and blocking.

Non-blocking write client

Non-blocking write client uses implicit batching. Data are asynchronously written to the underlying buffer and they are automatically sent to a server when the size of the write buffer reaches the batch size, default 5000, or the flush interval, default 1s, times out. Writes are automatically retried on server back pressure.

This write client also offers synchronous blocking method to ensure that write buffer is flushed and all pending writes are finished, see Flush() method. Always use Close() method of the client to stop all background processes.

Asynchronous write client is recommended for frequent periodic writes.

package main

import (
    "fmt"
    "math/rand"
    "time"
    
    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    // and set batch size to 20 
    client := influxdb2.NewClientWithOptions("http://localhost:8086", "my-token",
        influxdb2.DefaultOptions().SetBatchSize(20))
    // Get non-blocking write client
    writeAPI := client.WriteAPI("my-org","my-bucket")
    // write some points
    for i := 0; i <100; i++ {
        // create point
        p := influxdb2.NewPoint(
            "system",
            map[string]string{
                "id":       fmt.Sprintf("rack_%v", i%10),
                "vendor":   "AWS",
                "hostname": fmt.Sprintf("host_%v", i%100),
            },
            map[string]interface{}{
                "temperature": rand.Float64() * 80.0,
                "disk_free":   rand.Float64() * 1000.0,
                "disk_total":  (i/10 + 1) * 1000000,
                "mem_total":   (i/100 + 1) * 10000000,
                "mem_free":    rand.Uint64(),
            },
            time.Now())
        // write asynchronously
        writeAPI.WritePoint(p)
    }
    // Force all unwritten data to be sent
    writeAPI.Flush()
    // Ensures background processes finishes
    client.Close()
}

Handling of failed async writes

WriteAPI by default continues with retrying of failed writes. Retried are automatically writes that fail on a connection failure or when server returns response HTTP status code >= 429.

Retrying algorithm uses random exponential strategy to set retry time. The delay for the next retry attempt is a random value in the interval retryInterval * exponentialBase^(attempts) and retryInterval * exponentialBase^(attempts+1). If writes of batch repeatedly fails, WriteAPI continues with retrying until maxRetries is reached or the overall retry time of batch exceeds maxRetryTime.

The defaults parameters (part of the WriteOptions) are:

  • retryInterval=5,000ms
  • exponentialBase=2
  • maxRetryDelay=125,000ms
  • maxRetries=5
  • maxRetryTime=180,000ms

Retry delays are by default randomly distributed within the ranges:

  1. 5,000-10,000
  2. 10,000-20,000
  3. 20,000-40,000
  4. 40,000-80,000
  5. 80,000-125,000

Setting retryInterval to 0 disables retry strategy and any failed write will discard the batch.

WriteFailedCallback allows advanced controlling of retrying. It is synchronously notified in case async write fails. It controls further batch handling by its return value. If it returns true, WriteAPI continues with retrying of writes of this batch. Returned false means the batch should be discarded.

Reading async errors

Errors() method returns a channel for reading errors which occurs during async writes. This channel is unbuffered and it must be read asynchronously otherwise will block write procedure:

package main

import (
    "fmt"
    "math/rand"
    "time"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Get non-blocking write client
    writeAPI := client.WriteAPI("my-org", "my-bucket")
    // Get errors channel
    errorsCh := writeAPI.Errors()
    // Create go proc for reading and logging errors
    go func() {
        for err := range errorsCh {
            fmt.Printf("write error: %s\n", err.Error())
        }
    }()
    // write some points
    for i := 0; i < 100; i++ {
        // create point
        p := influxdb2.NewPointWithMeasurement("stat").
            AddTag("id", fmt.Sprintf("rack_%v", i%10)).
            AddTag("vendor", "AWS").
            AddTag("hostname", fmt.Sprintf("host_%v", i%100)).
            AddField("temperature", rand.Float64()*80.0).
            AddField("disk_free", rand.Float64()*1000.0).
            AddField("disk_total", (i/10+1)*1000000).
            AddField("mem_total", (i/100+1)*10000000).
            AddField("mem_free", rand.Uint64()).
            SetTime(time.Now())
        // write asynchronously
        writeAPI.WritePoint(p)
    }
    // Force all unwritten data to be sent
    writeAPI.Flush()
    // Ensures background processes finishes
    client.Close()
}

Blocking write client

Blocking write client writes given point(s) synchronously. It doesn't have implicit batching. Batch is created from given set of points.

package main

import (
    "context"
    "fmt"
    "math/rand"
    "time"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Get blocking write client
    writeAPI := client.WriteAPIBlocking("my-org","my-bucket")
    // write some points
    for i := 0; i <100; i++ {
        // create data point
        p := influxdb2.NewPoint(
            "system",
            map[string]string{
                "id":       fmt.Sprintf("rack_%v", i%10),
                "vendor":   "AWS",
                "hostname": fmt.Sprintf("host_%v", i%100),
            },
            map[string]interface{}{
                "temperature": rand.Float64() * 80.0,
                "disk_free":   rand.Float64() * 1000.0,
                "disk_total":  (i/10 + 1) * 1000000,
                "mem_total":   (i/100 + 1) * 10000000,
                "mem_free":    rand.Uint64(),
            },
            time.Now())
        // write synchronously
        err := writeAPI.WritePoint(context.Background(), p)
        if err != nil {
            panic(err)
        }
    }
    // Ensures background processes finishes
    client.Close()
}

Queries

Query client offers retrieving of query results to a parsed representation in a QueryTableResult or to a raw string.

QueryTableResult

QueryTableResult offers comfortable way how to deal with flux query CSV response. It parses CSV stream into FluxTableMetaData, FluxColumn and FluxRecord objects for easy reading the result.

package main

import (
    "context"
    "fmt"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Get query client
    queryAPI := client.QueryAPI("my-org")
    // get QueryTableResult
    result, err := queryAPI.Query(context.Background(), `from(bucket:"my-bucket")|> range(start: -1h) |> filter(fn: (r) => r._measurement == "stat")`)
    if err == nil {
        // Iterate over query response
        for result.Next() {
            // Notice when group key has changed
            if result.TableChanged() {
                fmt.Printf("table: %s\n", result.TableMetadata().String())
            }
            // Access data
            fmt.Printf("value: %v\n", result.Record().Value())
        }
        // check for an error
        if result.Err() != nil {
            fmt.Printf("query parsing error: %s\n", result.Err().Error())
        }
    } else {
        panic(err)
    }
    // Ensures background processes finishes
    client.Close()
}

Raw

QueryRaw() returns raw, unparsed, query result string and process it on your own. Returned csv format
can be controlled by the third parameter, query dialect.

package main

import (
    "context"
    "fmt"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    // Create a new client using an InfluxDB server base URL and an authentication token
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Get query client
    queryAPI := client.QueryAPI("my-org")
    // Query and get complete result as a string
    // Use default dialect
    result, err := queryAPI.QueryRaw(context.Background(), `from(bucket:"my-bucket")|> range(start: -1h) |> filter(fn: (r) => r._measurement == "stat")`, influxdb2.DefaultDialect())
    if err == nil {
        fmt.Println("QueryResult:")
        fmt.Println(result)
    } else {
        panic(err)
    }
    // Ensures background processes finishes
    client.Close()
}    

Concurrency

InfluxDB Go Client can be used in a concurrent environment. All its functions are thread-safe.

The best practise is to use a single Client instance per server URL. This ensures optimized resources usage, most importantly reusing HTTP connections.

For efficient reuse of HTTP resources among multiple clients, create an HTTP client and use Options.SetHTTPClient() for setting it to all clients:

    // Create HTTP client
    httpClient := &http.Client{
        Timeout: time.Second * time.Duration(60),
        Transport: &http.Transport{
            DialContext: (&net.Dialer{
                Timeout: 5 * time.Second,
            }).DialContext,
            TLSHandshakeTimeout: 5 * time.Second,
            TLSClientConfig: &tls.Config{
                InsecureSkipVerify: true,
            },
            MaxIdleConns:        100,
            MaxIdleConnsPerHost: 100,
            IdleConnTimeout:     90 * time.Second,
        },
    }
    // Client for server 1
    client1 := influxdb2.NewClientWithOptions("https://server:8086", "my-token", influxdb2.DefaultOptions().SetHTTPClient(httpClient))
    // Client for server 2
    client2 := influxdb2.NewClientWithOptions("https://server:9999", "my-token2", influxdb2.DefaultOptions().SetHTTPClient(httpClient))

Client ensures that there is a single instance of each server API sub-client for the specific area. E.g. a single WriteAPI instance for each org/bucket pair, a single QueryAPI for each org.

Such a single API sub-client instance can be used concurrently:

package main

import (
	"math/rand"
	"sync"
	"time"

	influxdb2 "github.com/influxdata/influxdb-client-go"
	"github.com/influxdata/influxdb-client-go/v2/api/write"
)

func main() {
    // Create client
    client := influxdb2.NewClient("http://localhost:8086", "my-token")
    // Ensure closing the client
    defer client.Close()

    // Get write client
    writeApi := client.WriteAPI("my-org", "my-bucket")

    // Create channel for points feeding
    pointsCh := make(chan *write.Point, 200)

    threads := 5

    var wg sync.WaitGroup
    go func(points int) {
        for i := 0; i < points; i++ {
            p := influxdb2.NewPoint("meas",
                map[string]string{"tag": "tagvalue"},
                map[string]interface{}{"val1": rand.Int63n(1000), "val2": rand.Float64()*100.0 - 50.0},
                time.Now())
            pointsCh <- p
        }
        close(pointsCh)
    }(1000000)

    // Launch write routines
    for t := 0; t < threads; t++ {
        wg.Add(1)
        go func() {
            for p := range pointsCh {
                writeApi.WritePoint(p)
            }
            wg.Done()
        }()
    }
    // Wait for writes complete
    wg.Wait()
}

Proxy and redirects

You can configure InfluxDB Go client behind a proxy in two ways:

  1. Using environment variable
    Set environment variable HTTP_PROXY (or HTTPS_PROXY based on the scheme of your server url).
    e.g. (linux) export HTTP_PROXY=http://my-proxy:8080 or in Go code os.Setenv("HTTP_PROXY","http://my-proxy:8080")

  2. Configure http.Client to use proxy
    Create a custom http.Client with a proxy configuration:

    proxyUrl, err := url.Parse("http://my-proxy:8080")
    httpClient := &http.Client{
        Transport: &http.Transport{
            Proxy: http.ProxyURL(proxyUrl)
        }
    }
    client := influxdb2.NewClientWithOptions("http://localhost:8086", token, influxdb2.DefaultOptions().SetHTTPClient(httpClient))

Client automatically follows HTTP redirects. The default redirect policy is to follow up to 10 consecutive requests. Due to a security reason Authorization header is not forwarded when redirect leads to a different domain. To overcome this limitation you have to set a custom redirect handler:

token := "my-token"

httpClient := &http.Client{
    CheckRedirect: func(req *http.Request, via []*http.Request) error {
        req.Header.Add("Authorization","Token " + token)
        return nil
    },
}
client := influxdb2.NewClientWithOptions("http://localhost:8086", token, influxdb2.DefaultOptions().SetHTTPClient(httpClient))

Checking Server State

There are three functions for checking whether a server is up and ready for communication:

Function Description Availability
Health() Detailed info about the server status, along with version string OSS
Ready() Server uptime info OSS
Ping() Whether a server is up OSS, Cloud

Only the Ping() function works in InfluxDB Cloud server.

InfluxDB 1.8 API compatibility

InfluxDB 1.8.0 introduced forward compatibility APIs for InfluxDB 2.0. This allow you to easily move from InfluxDB 1.x to InfluxDB 2.0 Cloud or open source.

Client API usage differences summary:

  1. Use the form username:password for an authentication token. Example: my-user:my-password. Use an empty string ("") if the server doesn't require authentication.
  2. The organization parameter is not used. Use an empty string ("") where necessary.
  3. Use the form database/retention-policy where a bucket is required. Skip retention policy if the default retention policy should be used. Examples: telegraf/autogen, telegraf.  

The following forward compatible APIs are available:

API Endpoint Description
WriteAPI (also WriteAPIBlocking) /api/v2/write Write data to InfluxDB 1.8.0+ using the InfluxDB 2.0 API
QueryAPI /api/v2/query Query data in InfluxDB 1.8.0+ using the InfluxDB 2.0 API and Flux endpoint should be enabled by the flux-enabled option
Health() /health Check the health of your InfluxDB instance

Example

package main

import (
    "context"
    "fmt"
    "time"

    "github.com/influxdata/influxdb-client-go/v2"
)

func main() {
    userName := "my-user"
    password := "my-password"
     // Create a new client using an InfluxDB server base URL and an authentication token
    // For authentication token supply a string in the form: "username:password" as a token. Set empty value for an unauthenticated server
    client := influxdb2.NewClient("http://localhost:8086", fmt.Sprintf("%s:%s",userName, password))
    // Get the blocking write client
    // Supply a string in the form database/retention-policy as a bucket. Skip retention policy for the default one, use just a database name (without the slash character)
    // Org name is not used
    writeAPI := client.WriteAPIBlocking("", "test/autogen")
    // create point using full params constructor
    p := influxdb2.NewPoint("stat",
        map[string]string{"unit": "temperature"},
        map[string]interface{}{"avg": 24.5, "max": 45},
        time.Now())
    // Write data
    err := writeAPI.WritePoint(context.Background(), p)
    if err != nil {
        fmt.Printf("Write error: %s\n", err.Error())
    }

    // Get query client. Org name is not used
    queryAPI := client.QueryAPI("")
    // Supply string in a form database/retention-policy as a bucket. Skip retention policy for the default one, use just a database name (without the slash character)
    result, err := queryAPI.Query(context.Background(), `from(bucket:"test")|> range(start: -1h) |> filter(fn: (r) => r._measurement == "stat")`)
    if err == nil {
        for result.Next() {
            if result.TableChanged() {
                fmt.Printf("table: %s\n", result.TableMetadata().String())
            }
            fmt.Printf("row: %s\n", result.Record().String())
        }
        if result.Err() != nil {
            fmt.Printf("Query error: %s\n", result.Err().Error())
        }
    } else {
        fmt.Printf("Query error: %s\n", err.Error())
    }
    // Close client
    client.Close()
}

Contributing

If you would like to contribute code you can do through GitHub by forking the repository and sending a pull request into the master branch.

License

The InfluxDB 2 Go Client is released under the MIT License.

About

InfluxDB 2 Go Client

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Go 99.1%
  • Shell 0.9%