InfluxDB Plugin for Jenkins

Description

Collects data from various other Jenkins plugins and sends metrics to InfluxDB. It's also possible to send custom data inside pipeline jobs.

Breaking Changes

Support for InfluxDB 1.7 and lower was dropped in plugin release 3.0.

You can find all breaking changes for each version in the breaking changes documentation.

Configuration

Configuration as Code

To configure InfluxDB plugin in Jenkins add the following to jenkins.yaml:

unclassified:
  influxDbGlobalConfig:
    targets:
    - credentialsId: "some_id"
      database: "some_database"
      description: "some description"
      exposeExceptions: true
      globalListener: true
      globalListenerFilter: "some filter"
      jobScheduledTimeAsPointsTimestamp: true
      retentionPolicy: "some_policy"
      url: "http://some/url"
      usingJenkinsProxy: true

⚠️ Prior 3.0, credentialsId needs to swapped to username and password.

Via Jenkins UI

Create a database in InfluxDB and a user with access rights. In Jenkins, go to Manage Jenkins > Configure System > InfluxDB Targets and click "Add". Provide the database information. The "URL" parameter requires the whole URL of the InfluxDB database, including the http(s):// and the database port. Also, provide the retention policy you want the data to be stored in InfluxDB (e.g. 15m or 2d). By default, it is infinite. Exceptions generated by the InfluxDB plugin can also be ignored by deselecting the "Expose Exceptions" checkbox.

⚠️ When you are using InfluxDB 1.x targets, leave the Organization blank. InfluxDB 2.x uses organizations and buckets instead of databases and retention policies, so organization field is used as an identifier whether the target is InfluxDB 1.x or 2.x.

In your job, select "Publish build data to InfluxDB" from the post-build actions.

Via Jenkins Pipeline

From version 1.19 onwards, you can create and remove targets in pipelines directly.

// Get InfluxDB plugin descriptor (version < 2.0)
def influxdb = Jenkins.instance.getDescriptorByType(jenkinsci.plugins.influxdb.DescriptorImpl)

// version >= 2.0
def influxdb = Jenkins.instance.getDescriptorByType(jenkinsci.plugins.influxdb.InfluxDbStep.DescriptorImpl)

// Create target
def target = new jenkinsci.plugins.influxdb.models.Target()

// Set target details

// Mandatory fields
target.description = 'my-new-target'
target.url = 'http://influxdburl:8086'

// Not supported >= 3.0
target.username = 'my-username'

// version < 2.0
target.password = 'my-password'

// version 2.0 to 2.6
target.password = hudson.util.Secret.fromString('my-password')

// Version >= 3.0
target.credentialsId = 'my-id'

target.database = 'my-database'

// Optional fields
target.retentionPolicy = '1d'                    // default = 'autogen'
target.jobScheduledTimeAsPointsTimestamp = true  // default = false
target.exposeExceptions = true                   // default = true
target.usingJenkinsProxy = true                  // default = false

// Add a target by using the created target object
influxdb.addTarget(target)
influxdb.save()

// Write stuff to InfluxDB
influxDbPublisher(selectedTarget: 'my-new-target')

// Remove a target by using the target description field value
influxdb.removeTarget('my-new-target')
influxdb.save()

Credentials

InfluxDB 1.8 targets support only Username with password.
InfluxDB 2.x supports Username with password for basic authentication and Secret Text for authentication token authentication.

Usage

Global Listener

When globalListener is set to true for a target in which no results were published during the build, it will automatically publish the result for this target when the build is completed.

To configure the global listener, you can use environment variables prefixed with INFLUXDB_PLUGIN. The following variables are supported and all correspond to an influxDbPublisher optional parameter.

INFLUXDB_PLUGIN_CUSTOM_PROJECT_NAME -> customProjectName
INFLUXDB_PLUGIN_CUSTOM_PREFIX -> customPrefix
INFLUXDB_PLUGIN_CUSTOM_FIELDS -> jenkinsEnvParameterField
INFLUXDB_PLUGIN_CUSTOM_TAGS -> jenkinsEnvParameterTag

NOTE: The environment variables must be set on the final build object. If you are creating or updating these variables in a pipeline, you should make sure they are exported with an EnvironmentContributingAction.

Freestyle Jobs

Select the InfluxDB target you wish to publish the data to.

From the "Advanced" tab you can choose to set a custom prefix for your project_name field, a custom project name to be used instead of the default job name, custom fields for your jenkins_data metric, and custom tags for your all your metrics.

Pipelines

The plugin can be used by calling either the influxDbPublisher() or the step() function.

NOTE: The influxDbPublisher() function is only supported from version 1.21 onwards.

Pipeline syntax

The only mandatory parameter is selectedTarget, which is the "Description" for your target in the global configuration.

influxDbPublisher(selectedTarget: 'my-target')

step([$class: 'InfluxDbPublisher', selectedTarget: 'my-target'])

Optional parameters

customProjectName (String) - custom project name
customPrefix (String) - custom prefix for project name
customData (Map) - custom fields in "jenkins_custom_data" measurement
customDataTags (Map) - custom tags in "jenkins_custom_data" measurement
customDataMap (Map) - custom fields in custom measurements
customDataMapTags (Map) - custom tags in custom measurements (map of tags for each custom measurements)
jenkinsEnvParameterField (String) - custom fields in "jenkins_data" measurement (newline-separated KEY=VALUE pairs)
jenkinsEnvParameterTag (String) - custom tags in all measurements (newline-separated KEY=VALUE pairs)
measurementName (String) - custom measurement name (replaces default "jenkins_data" and "jenkins_custom_data")

All customData* parameters contain custom data generated during the build and not by the plugin, so they are not available in the snippet generator.

❗ NOTE! Up to release 1.10.3, pipeline was configured with using the url and database.
step([$class: 'InfluxDbPublisher',
       target: 'http://127.0.0.1:8086,jenkins_db',
       ...
   ])
This form of configuration is not supported from version 1.11 onwards.

Pipelines don't have post-build actions, so the build result, build ordinal, and the build success boolean will default to "?", 5, and false respectively, unless set manually before calling InfluxDbPublisher. Only the build result needs to be set manually, as the boolean value and ordinal are set based on build result. Also, the build status will appear as "?" and the build duration might be a little off, because the build is not actually finished. If you want to get those pieces of information you need to configure the plugin separately on each job as a post-build action. The jobs can be run with, for example, the Build Pipeline Plugin to get data from all jobs to InfluxDB. Alternatively, you can insert the information in your build manually inside your Groovy script.

try {
    // Build things here
    if (currentBuild.result == null) {
        currentBuild.result = 'SUCCESS' // sets the ordinal as 0 and boolean to true
    }
} catch (err) {
    if (currentBuild.result == null) {
        currentBuild.result = 'FAILURE' // sets the ordinal as 4 and boolean to false
    }
    throw err
} finally {
    influxDbPublisher(selectedTarget: 'my-target')
}

Custom Data

You can write custom data to InfluxDB like this:

def myFields = [:]
myFields['field_a'] = 11
myFields['field_b'] = 12
influxDbPublisher(selectedTarget: 'my-target', customData: myFields)

This adds the fields field_a and field_b with values 11 and 12 respectively to a measurement called jenkins_custom_data.

You can also add tags to this measurement with the customDataTags parameter.

Alternatively, you can write custom data to InfluxDB with a higher degree of customization like this:

def myFields1 = [:]
def myFields2 = [:]
def myCustomMeasurementFields = [:]
myFields1['field_a'] = 11
myFields1['field_b'] = 12
myFields2['field_c'] = 21
myFields2['field_d'] = 22
myCustomMeasurementFields['series_1'] = myFields1
myCustomMeasurementFields['series_2'] = myFields2
myTags = ['series_1':['tag_a':'a','tag_b':'b'],'series_2':['tag_c':'c','tag_d':'d']]
influxDbPublisher(selectedTarget: 'my-target', customDataMap: myCustomMeasurementFields, customDataMapTags: myTags)

This creates 2 measurements, series_1 and series_2. It adds the fields field_a and field_b with values 11 and 12 respectively to measurement series_1. It adds the fields field_c and field_d with values 21 and 22 respectively to measurement series_2.

You can also add tags to your custom measurements with the customDataMapTags parameter. You must use the same map keys as measurement names as in customDataMap.

Supported Metrics

See all available metrics from the available metrics documentation.

Contribution

Create a pull request to the development branch. No pull requests are merged directly to master. Comment your changes sufficiently and create appropriate tests.

For feature requests and bug reports, please use the Jenkins issue tracker.

Acknowledgements

This plugin was inspired by Jouni Rajala and Christoph Burmeister.

groupe-edf/influxdb-plugin