Create a task

Create tasks with the InfluxDB user interface (UI), influx command line interface (CLI), or /api/v2 API.

Before creating a task, review the basics for writing a task.

• InfluxDB UI
• influx CLI
• InfluxDB API

Create a task in the InfluxDB UI

The InfluxDB UI provides multiple ways to create a task:

• Create a task from the Data Explorer
• Create a task in the Task UI
• Import a task
• Create a task from a template
• Clone a task

Create a task from the Data Explorer

1. In the navigation menu on the left, select Data Explorer.

   Data Explorer

2. Build a query and click Save As in the upper right.

3. Select the Task heading.

4. Specify the task options. See Task options for detailed information about each option.

5. Click Save as Task.

Create a task in the Task UI

1. In the navigation menu on the left, select Tasks.

   Tasks

2. Click Create Task in the upper right.

3. In the left panel, specify the task options. See Task options for detailed information about each option.

4. In the right panel, enter your task script.

   Leave out the option tasks assignment

   When creating a new task in the InfluxDB Task UI, leave the code editor empty. When you save the task, the Task UI uses the task options you specify in the Task options form to populate option task = {task_options} for you.

   When you edit the saved task, you’ll see the injected option task = {task_options}.

5. Click Save in the upper right.

Import a task

1. In the navigation menu on the left, select Tasks.

   Tasks

2. Click Create Task in the upper right.

3. In the left panel, specify the task options. See Task options for detailed information about each option.

4. Paste a raw Flux task in the code editor to the right of the task options fields.

5. Click Save in the upper right.

Create a task from a template

1. In the navigation menu on the left, select Settings > Templates.

   Settings

2. Find the template you want to use and click its Resources list to expand the list of resources.

3. In the Resources list, click the task you want to use.

Clone a task

1. In the navigation menu on the left, select Tasks.

   Tasks

2. Find the task you would like to clone and click the icon located far right of the task name.

3. Click Clone.

Create a task using the influx CLI

Use the influx task create command to create a new task. It accepts either a file path or raw Flux.

Create a task using a file

# Syntax
influx task create --org <org-name>  -f </path/to/task-script>

# Example
influx task create --org my-org -f /tasks/cq-mean-1h.flux

Create a task using raw Flux

influx task create --org my-org - # <return> to open stdin pipe

option task = {
  name: "task-name",
  every: 6h
}

# ... Task script ...

# Linux & macOS: <ctrl-d> to close the pipe and submit the command
# Windows: <enter>, then <ctrl-d>, then <enter> to close the pipe and submit the command

Create a task using the InfluxDB API

POST http://localhost:8086/api/v2/tasks/

curl --request POST 'https://cloud2.influxdata.com/api/v2/tasks' \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Token INFLUX_API_TOKEN' \
    "cron": "0 * * * *",
    "name": "downsample cpu",
    "scriptID": "085a2960eaa20000",
    "description": "This task downsamples CPU data every hour"
}'

SCRIPT_ID=$(
curl https://cloud2.influxdata.com/api/v2/scripts \
  --header "Authorization: Token INFLUX_API_TOKEN" \
  --header 'Accept: application/json' \
  --header 'Content-Type: application/json' \
  --data-binary @- << EOF | jq -r '.id'
  {
    "name": "filter-and-group19",
    "description": "Returns filtered and grouped points from a bucket.",
    "script": "from(bucket: params.bucket)\
               |> range(start: duration(v: params.rangeStart))\
               |> filter(fn: (r) => r._field == params.filterField)\
               |> group(columns: [params.groupColumn])",
     "language": "flux"
  }
EOF
)

echo $SCRIPT_ID

curl https://cloud2.influxdata.com/api/v2/tasks \
--header "Content-type: application/json" \
--header "Authorization: Token INFLUX_API_TOKEN" \
--data @- << EOF
  {
  "name": "30-day-avg-temp",
  "description": "IoT Center 30d temperature average.",
  "every": "1d",
  "scriptID": "${SCRIPT_ID}",
  "scriptParameters":
    {
      "rangeStart": "-30d",
      "bucket": "air_sensor",
      "filterField": "temperature",
      "groupColumn": "_time"
    }
  }
EOF

POST https://cloud2.influxdata.com/api/v2/tasks/

curl --request POST 'https://cloud2.influxdata.com/api/v2/tasks' \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Token INFLUX_API_TOKEN' \
    --data-binary @- << EOF
      {
        "flux": "option task = {name: \"CPU Total 1 Hour New\", every: 1h}\
        from(bucket: \"telegraf\")
          |> range(start: -1h)
          |> filter(fn: (r) => (r._measurement == \"cpu\"))
          |> filter(fn: (r) =>\n\t\t(r._field == \"usage_system\"))
          |> filter(fn: (r) => (r.cpu == \"cpu-total\"))
          |> aggregateWindow(every: 1h, fn: max)
          |> to(bucket: \"cpu_usage_user_total_1h\", org: \"INFLUX_ORG\")",
        "orgID": "INFLUX_ORG_ID",
        "status": "active",
        "description": "This task downsamples CPU data every hour"
      }
EOF