search

influxdata / influxdb-client-python

InfluxDB 2.0 python client
https://influxdb-client.readthedocs.io/en/stable/
MIT License
724 stars 187 forks source link
flux influxdata influxdb jupyter pandas-dataframe python reactive timeseries

readme

influxdb-client-python

CircleCI codecov CI status PyPI package Anaconda.org package Supported Python versions Documentation status Slack Status

This repository contains the Python client library for use with InfluxDB 2.x and Flux. InfluxDB 3.x users should instead use the lightweight v3 client library. InfluxDB 1.x users should use the v1 client library.

For ease of migration and a consistent query and write experience, v2 users should consider using InfluxQL and the v1 client library.

The API of the influxdb-client-python is not the backwards-compatible with the old one - influxdb-python.

Documentation

This section contains links to the client library documentation.

InfluxDB 2.0 client features

Installation

InfluxDB python library uses RxPY - The Reactive Extensions for Python (RxPY).

Python 3.7 or later is required.

:warning:

It is recommended to use ciso8601 with client for parsing dates. ciso8601 is much faster than built-in Python datetime. Since it's written as a C module the best way is build it from sources:

Windows:

You have to install Visual C++ Build Tools 2015 to build ciso8601 by pip.

conda:

Install from sources: conda install -c conda-forge/label/cf202003 ciso8601.

pip install

The python package is hosted on PyPI, you can install latest version directly:

pip install 'influxdb-client[ciso]'

Then import the package:

import influxdb_client

If your application uses async/await in Python you can install with the async extra:

$ pip install influxdb-client[async]

For more info see How to use Asyncio.

Setuptools

Install via Setuptools.

python setup.py install --user

(or sudo python setup.py install to install the package for all users)

Getting Started

Please follow the Installation and then run the following:

from influxdb_client import InfluxDBClient, Point
from influxdb_client.client.write_api import SYNCHRONOUS

bucket = "my-bucket"

client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org")

write_api = client.write_api(write_options=SYNCHRONOUS)
query_api = client.query_api()

p = Point("my_measurement").tag("location", "Prague").field("temperature", 25.3)

write_api.write(bucket=bucket, record=p)

## using Table structure
tables = query_api.query('from(bucket:"my-bucket") |> range(start: -10m)')

for table in tables:
    print(table)
    for row in table.records:
        print (row.values)

## using csv library
csv_result = query_api.query_csv('from(bucket:"my-bucket") |> range(start: -10m)')
val_count = 0
for row in csv_result:
    for cell in row:
        val_count += 1

Client configuration

Via File

A client can be configured via *.ini file in segment influx2.

The following options are supported:

self.client = InfluxDBClient.from_config_file("config.ini")
[influx2]
url=http://localhost:8086
org=my-org
token=my-token
timeout=6000
verify_ssl=False

Via Environment Properties

A client can be configured via environment properties.

Supported properties are:

self.client = InfluxDBClient.from_env_properties()

Profile query

The Flux Profiler package provides performance profiling tools for Flux queries and operations.

You can enable printing profiler information of the Flux query in client library by:

When the profiler is enabled, the result of flux query contains additional tables "profiler/". In order to have consistent behaviour with enabled/disabled profiler, FluxCSVParser excludes "profiler/" measurements from result.

Example how to enable profilers using API:

q = '''
    from(bucket: stringParam)
      |> range(start: -5m, stop: now())
      |> filter(fn: (r) => r._measurement == "mem")
      |> filter(fn: (r) => r._field == "available" or r._field == "free" or r._field == "used")
      |> aggregateWindow(every: 1m, fn: mean)
      |> pivot(rowKey:["_time"], columnKey: ["_field"], valueColumn: "_value")
'''
p = {
    "stringParam": "my-bucket",
}

query_api = client.query_api(query_options=QueryOptions(profilers=["query", "operator"]))
csv_result = query_api.query(query=q, params=p)

Example of a profiler output:

===============
Profiler: query
===============

from(bucket: stringParam)
  |> range(start: -5m, stop: now())
  |> filter(fn: (r) => r._measurement == "mem")
  |> filter(fn: (r) => r._field == "available" or r._field == "free" or r._field == "used")
  |> aggregateWindow(every: 1m, fn: mean)
  |> pivot(rowKey:["_time"], columnKey: ["_field"], valueColumn: "_value")

========================
Profiler: profiler/query
========================
result              : _profiler
table               : 0
_measurement        : profiler/query
TotalDuration       : 8924700
CompileDuration     : 350900
QueueDuration       : 33800
PlanDuration        : 0
RequeueDuration     : 0
ExecuteDuration     : 8486500
Concurrency         : 0
MaxAllocated        : 2072
TotalAllocated      : 0
flux/query-plan     :

digraph {
  ReadWindowAggregateByTime11
  // every = 1m, aggregates = [mean], createEmpty = true, timeColumn = "_stop"
  pivot8
  generated_yield

  ReadWindowAggregateByTime11 -> pivot8
  pivot8 -> generated_yield
}

influxdb/scanned-bytes: 0
influxdb/scanned-values: 0

===========================
Profiler: profiler/operator
===========================
result              : _profiler
table               : 1
_measurement        : profiler/operator
Type                : *universe.pivotTransformation
Label               : pivot8
Count               : 3
MinDuration         : 32600
MaxDuration         : 126200
DurationSum         : 193400
MeanDuration        : 64466.666666666664

===========================
Profiler: profiler/operator
===========================
result              : _profiler
table               : 1
_measurement        : profiler/operator
Type                : *influxdb.readWindowAggregateSource
Label               : ReadWindowAggregateByTime11
Count               : 1
MinDuration         : 940500
MaxDuration         : 940500
DurationSum         : 940500
MeanDuration        : 940500.0

You can also use callback function to get profilers output. Return value of this callback is type of FluxRecord.

Example how to use profilers with callback:

class ProfilersCallback(object):
   def __init__(self):
       self.records = []

   def __call__(self, flux_record):
       self.records.append(flux_record.values)

callback = ProfilersCallback()

query_api = client.query_api(query_options=QueryOptions(profilers=["query", "operator"], profiler_callback=callback))
tables = query_api.query('from(bucket:"my-bucket") |> range(start: -10m)')

for profiler in callback.records:
   print(f'Custom processing of profiler result: {profiler}')

Example output of this callback:

Custom processing of profiler result: {'result': '_profiler', 'table': 0, '_measurement': 'profiler/query', 'TotalDuration': 18843792, 'CompileDuration': 1078666, 'QueueDuration': 93375, 'PlanDuration': 0, 'RequeueDuration': 0, 'ExecuteDuration': 17371000, 'Concurrency': 0, 'MaxAllocated': 448, 'TotalAllocated': 0, 'RuntimeErrors': None, 'flux/query-plan': 'digraph {\r\n  ReadRange2\r\n  generated_yield\r\n\r\n  ReadRange2 -> generated_yield\r\n}\r\n\r\n', 'influxdb/scanned-bytes': 0, 'influxdb/scanned-values': 0}
Custom processing of profiler result: {'result': '_profiler', 'table': 1, '_measurement': 'profiler/operator', 'Type': '*influxdb.readFilterSource', 'Label': 'ReadRange2', 'Count': 1, 'MinDuration': 3274084, 'MaxDuration': 3274084, 'DurationSum': 3274084, 'MeanDuration': 3274084.0}

How to use

Writes

The WriteApi supports synchronous, asynchronous and batching writes into InfluxDB 2.0. The data should be passed as a InfluxDB Line Protocol, Data Point or Observable stream.

:warning:

The WriteApi in batching mode (default mode) is supposed to run as a singleton. To flush all your data you should wrap the execution using with client.write_api(...) as write_api: statement or call write_api.close() at the end of your script.

The default instance of WriteApi use batching.

The data could be written as

  1. string or bytes that is formatted as a InfluxDB's line protocol
  2. Data Point structure
  3. Dictionary style mapping with keys: measurement, tags, fields and time or custom structure
  4. NamedTuple
  5. Data Classes
  6. Pandas DataFrame
  7. List of above items
  8. A batching type of write also supports an Observable that produce one of an above item

You can find write examples at GitHub: influxdb-client-python/examples.

Batching

The batching is configurable by write_options:

Property Description Default Value
batch_size the number of data point to collect in a batch 1000
flush_interval the number of milliseconds before the batch is written 1000
jitter_interval the number of milliseconds to increase the batch flush interval by a random amount 0
retry_interval the number of milliseconds to retry first unsuccessful write. The next retry delay is computed using exponential random backoff. The retry interval is used when the InfluxDB server does not specify \"Retry-After\" header. 5000
max_retry_time maximum total retry timeout in milliseconds. 180_000
max_retries the number of max retries when write fails 5
max_retry_delay the maximum delay between each retry attempt in milliseconds 125_000
max_close_wait the maximum amount of time to wait for batches to flush when .close() is called 300_000
exponential_base the base for the exponential retry delay, the next delay is computed using random exponential backoff as a random value within the interval retry_interval * exponential_base^(attempts-1) and retry_interval * exponential_base^(attempts). Example for retry_interval=5_000, exponential_base=2, max_retry_delay=125_000, total=5 Retry delays are random distributed values within the ranges of [5_000-10_000, 10_000-20_000, 20_000-40_000, 40_000-80_000, 80_000-125_000] 2 
from datetime import datetime, timedelta, timezone

import pandas as pd
import reactivex as rx
from reactivex import operators as ops

from influxdb_client import InfluxDBClient, Point, WriteOptions

with InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org") as _client:

    with _client.write_api(write_options=WriteOptions(batch_size=500,
                                                      flush_interval=10_000,
                                                      jitter_interval=2_000,
                                                      retry_interval=5_000,
                                                      max_retries=5,
                                                      max_retry_delay=30_000,
                                                      max_close_wait=300_000,
                                                      exponential_base=2)) as _write_client:

        """
        Write Line Protocol formatted as string
        """
        _write_client.write("my-bucket", "my-org", "h2o_feet,location=coyote_creek water_level=1.0 1")
        _write_client.write("my-bucket", "my-org", ["h2o_feet,location=coyote_creek water_level=2.0 2",
                                                    "h2o_feet,location=coyote_creek water_level=3.0 3"])

        """
        Write Line Protocol formatted as byte array
        """
        _write_client.write("my-bucket", "my-org", "h2o_feet,location=coyote_creek water_level=1.0 1".encode())
        _write_client.write("my-bucket", "my-org", ["h2o_feet,location=coyote_creek water_level=2.0 2".encode(),
                                                    "h2o_feet,location=coyote_creek water_level=3.0 3".encode()])

        """
        Write Dictionary-style object
        """
        _write_client.write("my-bucket", "my-org", {"measurement": "h2o_feet", "tags": {"location": "coyote_creek"},
                                                    "fields": {"water_level": 1.0}, "time": 1})
        _write_client.write("my-bucket", "my-org", [{"measurement": "h2o_feet", "tags": {"location": "coyote_creek"},
                                                     "fields": {"water_level": 2.0}, "time": 2},
                                                    {"measurement": "h2o_feet", "tags": {"location": "coyote_creek"},
                                                     "fields": {"water_level": 3.0}, "time": 3}])

        """
        Write Data Point
        """
        _write_client.write("my-bucket", "my-org",
                            Point("h2o_feet").tag("location", "coyote_creek").field("water_level", 4.0).time(4))
        _write_client.write("my-bucket", "my-org",
                            [Point("h2o_feet").tag("location", "coyote_creek").field("water_level", 5.0).time(5),
                             Point("h2o_feet").tag("location", "coyote_creek").field("water_level", 6.0).time(6)])

        """
        Write Observable stream
        """
        _data = rx \
            .range(7, 11) \
            .pipe(ops.map(lambda i: "h2o_feet,location=coyote_creek water_level={0}.0 {0}".format(i)))

        _write_client.write("my-bucket", "my-org", _data)

        """
        Write Pandas DataFrame
        """
        _now = datetime.now(tz=timezone.utc)
        _data_frame = pd.DataFrame(data=[["coyote_creek", 1.0], ["coyote_creek", 2.0]],
                                   index=[_now, _now + timedelta(hours=1)],
                                   columns=["location", "water_level"])

        _write_client.write("my-bucket", "my-org", record=_data_frame, data_frame_measurement_name='h2o_feet',
                            data_frame_tag_columns=['location'])

Default Tags

Sometimes is useful to store same information in every measurement e.g. hostname, location, customer. The client is able to use static value or env property as a tag value.

The expressions:

Via API
point_settings = PointSettings()
point_settings.add_default_tag("id", "132-987-655")
point_settings.add_default_tag("customer", "California Miner")
point_settings.add_default_tag("data_center", "${env.data_center}")

self.write_client = self.client.write_api(write_options=SYNCHRONOUS, point_settings=point_settings)
self.write_client = self.client.write_api(write_options=SYNCHRONOUS,
                                              point_settings=PointSettings(**{"id": "132-987-655",
                                                                              "customer": "California Miner"}))
Via Configuration file

In an init configuration file you are able to specify default tags by tags segment.

self.client = InfluxDBClient.from_config_file("config.ini")
[influx2]
url=http://localhost:8086
org=my-org
token=my-token
timeout=6000

[tags]
id = 132-987-655
customer = California Miner
data_center = ${env.data_center}

You can also use a TOML or aJSON format for the configuration file.

Via Environment Properties

You are able to specify default tags by environment properties with prefix INFLUXDB_V2_TAG_.

Examples:

self.client = InfluxDBClient.from_env_properties()

Synchronous client

Data are writes in a synchronous HTTP request.

from influxdb_client import InfluxDBClient, Point
from influxdb_client .client.write_api import SYNCHRONOUS

client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org")
write_api = client.write_api(write_options=SYNCHRONOUS)

_point1 = Point("my_measurement").tag("location", "Prague").field("temperature", 25.3)
_point2 = Point("my_measurement").tag("location", "New York").field("temperature", 24.3)

write_api.write(bucket="my-bucket", record=[_point1, _point2])

client.close()

Queries

The result retrieved by QueryApi could be formatted as a:

  1. Flux data structure: FluxTable, FluxColumn and FluxRecord
  2. influxdb_client.client.flux_table.CSVIterator which will iterate over CSV lines
  3. Raw unprocessed results as a str iterator
  4. Pandas DataFrame

The API also support streaming FluxRecord via query_stream, see example below:

from influxdb_client import InfluxDBClient, Point, Dialect
from influxdb_client.client.write_api import SYNCHRONOUS

client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org")

write_api = client.write_api(write_options=SYNCHRONOUS)
query_api = client.query_api()

"""
Prepare data
"""

_point1 = Point("my_measurement").tag("location", "Prague").field("temperature", 25.3)
_point2 = Point("my_measurement").tag("location", "New York").field("temperature", 24.3)

write_api.write(bucket="my-bucket", record=[_point1, _point2])

"""
Query: using Table structure
"""
tables = query_api.query('from(bucket:"my-bucket") |> range(start: -10m)')

for table in tables:
    print(table)
    for record in table.records:
        print(record.values)

print()
print()

"""
Query: using Bind parameters
"""

p = {"_start": datetime.timedelta(hours=-1),
     "_location": "Prague",
     "_desc": True,
     "_floatParam": 25.1,
     "_every": datetime.timedelta(minutes=5)
     }

tables = query_api.query('''
    from(bucket:"my-bucket") |> range(start: _start)
        |> filter(fn: (r) => r["_measurement"] == "my_measurement")
        |> filter(fn: (r) => r["_field"] == "temperature")
        |> filter(fn: (r) => r["location"] == _location and r["_value"] > _floatParam)
        |> aggregateWindow(every: _every, fn: mean, createEmpty: true)
        |> sort(columns: ["_time"], desc: _desc)
''', params=p)

for table in tables:
    print(table)
    for record in table.records:
        print(str(record["_time"]) + " - " + record["location"] + ": " + str(record["_value"]))

print()
print()

"""
Query: using Stream
"""
records = query_api.query_stream('from(bucket:"my-bucket") |> range(start: -10m)')

for record in records:
    print(f'Temperature in {record["location"]} is {record["_value"]}')

"""
Interrupt a stream after retrieve a required data
"""
large_stream = query_api.query_stream('from(bucket:"my-bucket") |> range(start: -100d)')
for record in large_stream:
    if record["location"] == "New York":
        print(f'New York temperature: {record["_value"]}')
        break

large_stream.close()

print()
print()

"""
Query: using csv library
"""
csv_result = query_api.query_csv('from(bucket:"my-bucket") |> range(start: -10m)',
                                 dialect=Dialect(header=False, delimiter=",", comment_prefix="#", annotations=[],
                                                 date_time_format="RFC3339"))
for csv_line in csv_result:
    if not len(csv_line) == 0:
        print(f'Temperature in {csv_line[9]} is {csv_line[6]}')

"""
Close client
"""
client.close()

Pandas DataFrame

:warning:

For DataFrame querying you should install Pandas dependency via pip install 'influxdb-client[extra]'.

:warning:

Note that if a query returns more then one table than the client generates a DataFrame for each of them.

The client is able to retrieve data in Pandas DataFrame format thought query_data_frame:

from influxdb_client import InfluxDBClient, Point, Dialect
from influxdb_client.client.write_api import SYNCHRONOUS

client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org")

write_api = client.write_api(write_options=SYNCHRONOUS)
query_api = client.query_api()

"""
Prepare data
"""

_point1 = Point("my_measurement").tag("location", "Prague").field("temperature", 25.3)
_point2 = Point("my_measurement").tag("location", "New York").field("temperature", 24.3)

write_api.write(bucket="my-bucket", record=[_point1, _point2])

"""
Query: using Pandas DataFrame
"""
data_frame = query_api.query_data_frame('from(bucket:"my-bucket") '
                                        '|> range(start: -10m) '
                                        '|> pivot(rowKey:["_time"], columnKey: ["_field"], valueColumn: "_value") '
                                        '|> keep(columns: ["location", "temperature"])')
print(data_frame.to_string())

"""
Close client
"""
client.close()

Output:

result table  location  temperature
0  _result     0  New York         24.3
1  _result     1    Prague         25.3

Examples

How to efficiently import large dataset

The following example shows how to import dataset with a dozen megabytes. If you would like to import gigabytes of data then use our multiprocessing example: import_data_set_multiprocessing.py for use a full capability of your hardware.

"""
Import VIX - CBOE Volatility Index - from "vix-daily.csv" file into InfluxDB 2.0

https://datahub.io/core/finance-vix#data
"""

from collections import OrderedDict
from csv import DictReader

import reactivex as rx
from reactivex import operators as ops

from influxdb_client import InfluxDBClient, Point, WriteOptions

def parse_row(row: OrderedDict):
    """Parse row of CSV file into Point with structure:

        financial-analysis,type=ily close=18.47,high=19.82,low=18.28,open=19.82 1198195200000000000

    CSV format:
        Date,VIX Open,VIX High,VIX Low,VIX Close\n
        2004-01-02,17.96,18.68,17.54,18.22\n
        2004-01-05,18.45,18.49,17.44,17.49\n
        2004-01-06,17.66,17.67,16.19,16.73\n
        2004-01-07,16.72,16.75,15.5,15.5\n
        2004-01-08,15.42,15.68,15.32,15.61\n
        2004-01-09,16.15,16.88,15.57,16.75\n
        ...

    :param row: the row of CSV file
    :return: Parsed csv row to [Point]
    """

    """
     For better performance is sometimes useful directly create a LineProtocol to avoid unnecessary escaping overhead:
     """
     # from datetime import timezone
     # import ciso8601
     # from influxdb_client.client.write.point import EPOCH
     #
     # time = (ciso8601.parse_datetime(row["Date"]).replace(tzinfo=timezone.utc) - EPOCH).total_seconds() * 1e9
     # return f"financial-analysis,type=vix-daily" \
     #        f" close={float(row['VIX Close'])},high={float(row['VIX High'])},low={float(row['VIX Low'])},open={float(row['VIX Open'])} " \
     #        f" {int(time)}"

    return Point("financial-analysis") \
        .tag("type", "vix-daily") \
        .field("open", float(row['VIX Open'])) \
        .field("high", float(row['VIX High'])) \
        .field("low", float(row['VIX Low'])) \
        .field("close", float(row['VIX Close'])) \
        .time(row['Date'])

"""
Converts vix-daily.csv into sequence of datad point
"""
data = rx \
    .from_iterable(DictReader(open('vix-daily.csv', 'r'))) \
    .pipe(ops.map(lambda row: parse_row(row)))

client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org", debug=True)

"""
Create client that writes data in batches with 50_000 items.
"""
write_api = client.write_api(write_options=WriteOptions(batch_size=50_000, flush_interval=10_000))

"""
Write data into InfluxDB
"""
write_api.write(bucket="my-bucket", record=data)
write_api.close()

"""
Querying max value of CBOE Volatility Index
"""
query = 'from(bucket:"my-bucket")' \
        ' |> range(start: 0, stop: now())' \
        ' |> filter(fn: (r) => r._measurement == "financial-analysis")' \
        ' |> max()'
result = client.query_api().query(query=query)

"""
Processing results
"""
print()
print("=== results ===")
print()
for table in result:
    for record in table.records:
        print('max {0:5} = {1}'.format(record.get_field(), record.get_value()))

"""
Close client
"""
client.close()

Efficiency write data from IOT sensor

"""
Efficiency write data from IOT sensor - write changed temperature every minute
"""
import atexit
import platform
from datetime import timedelta

import psutil as psutil
import reactivex as rx
from reactivex import operators as ops

from influxdb_client import InfluxDBClient, WriteApi, WriteOptions

def on_exit(db_client: InfluxDBClient, write_api: WriteApi):
    """Close clients after terminate a script.

    :param db_client: InfluxDB client
    :param write_api: WriteApi
    :return: nothing
    """
    write_api.close()
    db_client.close()

def sensor_temperature():
    """Read a CPU temperature. The [psutil] doesn't support MacOS so we use [sysctl].

    :return: actual CPU temperature
    """
    os_name = platform.system()
    if os_name == 'Darwin':
        from subprocess import check_output
        output = check_output(["sysctl", "machdep.xcpm.cpu_thermal_level"])
        import re
        return re.findall(r'\d+', str(output))[0]
    else:
        return psutil.sensors_temperatures()["coretemp"][0]

def line_protocol(temperature):
    """Create a InfluxDB line protocol with structure:

        iot_sensor,hostname=mine_sensor_12,type=temperature value=68

    :param temperature: the sensor temperature
    :return: Line protocol to write into InfluxDB
    """

    import socket
    return 'iot_sensor,hostname={},type=temperature value={}'.format(socket.gethostname(), temperature)

"""
Read temperature every minute; distinct_until_changed - produce only if temperature change
"""
data = rx\
    .interval(period=timedelta(seconds=60))\
    .pipe(ops.map(lambda t: sensor_temperature()),
          ops.distinct_until_changed(),
          ops.map(lambda temperature: line_protocol(temperature)))

_db_client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org", debug=True)

"""
Create client that writes data into InfluxDB
"""
_write_api = _db_client.write_api(write_options=WriteOptions(batch_size=1))
_write_api.write(bucket="my-bucket", record=data)

"""
Call after terminate a script
"""
atexit.register(on_exit, _db_client, _write_api)

input()

Connect to InfluxDB Cloud

The following example demonstrate the simplest way how to write and query date with the InfluxDB Cloud.

At first point you should create an authentication token as is described here.

After that you should configure properties: influx_cloud_url,influx_cloud_token, bucket and org in a influx_cloud.py example.

The last step is run a python script via: python3 influx_cloud.py.

"""
Connect to InfluxDB 2.0 - write data and query them
"""

from datetime import datetime, timezone

from influxdb_client import Point, InfluxDBClient
from influxdb_client.client.write_api import SYNCHRONOUS

"""
Configure credentials
"""
influx_cloud_url = 'https://us-west-2-1.aws.cloud2.influxdata.com'
influx_cloud_token = '...'
bucket = '...'
org = '...'

client = InfluxDBClient(url=influx_cloud_url, token=influx_cloud_token)
try:
    kind = 'temperature'
    host = 'host1'
    device = 'opt-123'

    """
    Write data by Point structure
    """
    point = Point(kind).tag('host', host).tag('device', device).field('value', 25.3).time(time=datetime.now(tz=timezone.utc))

    print(f'Writing to InfluxDB cloud: {point.to_line_protocol()} ...')

    write_api = client.write_api(write_options=SYNCHRONOUS)
    write_api.write(bucket=bucket, org=org, record=point)

    print()
    print('success')
    print()
    print()

    """
    Query written data
    """
    query = f'from(bucket: "{bucket}") |> range(start: -1d) |> filter(fn: (r) => r._measurement == "{kind}")'
    print(f'Querying from InfluxDB cloud: "{query}" ...')
    print()

    query_api = client.query_api()
    tables = query_api.query(query=query, org=org)

    for table in tables:
        for row in table.records:
            print(f'{row.values["_time"]}: host={row.values["host"]},device={row.values["device"]} '
                  f'{row.values["_value"]} °C')

    print()
    print('success')

except Exception as e:
    print(e)
finally:
    client.close()

How to use Jupyter + Pandas + InfluxDB 2

The first example shows how to use client capabilities to predict stock price via Keras, TensorFlow, sklearn:

The example is taken from Kaggle.

image

Result:

image

The second example shows how to use client capabilities to realtime visualization via hvPlot, Streamz, RxPY:

image

Other examples

You can find all examples at GitHub: influxdb-client-python/examples.

Advanced Usage

Gzip support

InfluxDBClient does not enable gzip compression for http requests by default. If you want to enable gzip to reduce transfer data's size, you can call:

from influxdb_client import InfluxDBClient

_db_client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org", enable_gzip=True)

Authenticate to the InfluxDB

InfluxDBClient supports three options how to authorize a connection:

Token

Use the token to authenticate to the InfluxDB API. In your API requests, an Authorization header will be sent. The header value, provide the word Token followed by a space and an InfluxDB API token. The word token is case-sensitive.

from influxdb_client import InfluxDBClient

with InfluxDBClient(url="http://localhost:8086", token="my-token") as client

:warning:

Note that this is a preferred way how to authenticate to InfluxDB API.

Username & Password

Authenticates via username and password credentials. If successful, creates a new session for the user.

from influxdb_client import InfluxDBClient

with InfluxDBClient(url="http://localhost:8086", username="my-user", password="my-password") as client

:warning:

The username/password auth is based on the HTTP "Basic" authentication. The authorization expires when the time-to-live (TTL) (default 60 minutes) is reached and client produces unauthorized exception.

HTTP Basic

Use this to enable basic authentication when talking to a InfluxDB 1.8.x that does not use auth-enabled but is protected by a reverse proxy with basic authentication.

from influxdb_client import InfluxDBClient

with InfluxDBClient(url="http://localhost:8086", auth_basic=True, token="my-proxy-secret") as client

:warning:

Don't use this when directly talking to InfluxDB 2.

Proxy configuration

You can configure the client to tunnel requests through an HTTP proxy. The following proxy options are supported:

from influxdb_client import InfluxDBClient

with InfluxDBClient(url="http://localhost:8086",
                    token="my-token",
                    org="my-org",
                    proxy="http://localhost:3128") as client:

If your proxy notify the client with permanent redirect (HTTP 301) to different host. The client removes Authorization header, because otherwise the contents of Authorization is sent to third parties which is a security vulnerability.

You can change this behaviour by:

from urllib3 import Retry
Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT = frozenset()
Retry.DEFAULT.remove_headers_on_redirect = Retry.DEFAULT_REMOVE_HEADERS_ON_REDIRECT

Delete data

The delete_api.py supports deletes points from an InfluxDB bucket.

from influxdb_client import InfluxDBClient

client = InfluxDBClient(url="http://localhost:8086", token="my-token")

delete_api = client.delete_api()

"""
Delete Data
"""
start = "1970-01-01T00:00:00Z"
stop = "2021-02-01T00:00:00Z"
delete_api.delete(start, stop, '_measurement="my_measurement"', bucket='my-bucket', org='my-org')

"""
Close client
"""
client.close()

InfluxDB 1.8 API compatibility

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

The following forward compatible APIs are available:

API Endpoint Description
query_api.py /api/v2/query Query data in InfluxDB 1.8.0+ using the InfluxDB 2.0 API and Flux (endpoint should be enabled by flux-enabled option)
write_api.py /api/v2/write Write data to InfluxDB 1.8.0+ using the InfluxDB 2.0 API
ping() /ping Check the status of your InfluxDB instance

For detail info see InfluxDB 1.8 example.

Handling Errors

Errors happen, and it's important that your code is prepared for them. All client related exceptions are delivered from InfluxDBError. If the exception cannot be recovered in the client it is returned to the application. These exceptions are left for the developer to handle.

Almost all APIs directly return unrecoverable exceptions to be handled this way:

from influxdb_client import InfluxDBClient
from influxdb_client.client.exceptions import InfluxDBError
from influxdb_client.client.write_api import SYNCHRONOUS

with InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org") as client:
    try:
        client.write_api(write_options=SYNCHRONOUS).write("my-bucket", record="mem,tag=a value=86")
    except InfluxDBError as e:
        if e.response.status == 401:
            raise Exception(f"Insufficient write permissions to 'my-bucket'.") from e
        raise

The only exception is batching WriteAPI (for more info see Batching) where you need to register custom callbacks to handle batch events. This is because this API runs in the background in a separate thread and isn't possible to directly return underlying exceptions.

from influxdb_client import InfluxDBClient
from influxdb_client.client.exceptions import InfluxDBError

class BatchingCallback(object):

    def success(self, conf: (str, str, str), data: str):
        print(f"Written batch: {conf}, data: {data}")

    def error(self, conf: (str, str, str), data: str, exception: InfluxDBError):
        print(f"Cannot write batch: {conf}, data: {data} due: {exception}")

    def retry(self, conf: (str, str, str), data: str, exception: InfluxDBError):
        print(f"Retryable error occurs for batch: {conf}, data: {data} retry: {exception}")

with InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org") as client:
    callback = BatchingCallback()
    with client.write_api(success_callback=callback.success,
                          error_callback=callback.error,
                          retry_callback=callback.retry) as write_api:
        pass

HTTP Retry Strategy

By default, the client uses a retry strategy only for batching writes (for more info see Batching). For other HTTP requests there is no one retry strategy, but it could be configured by retries parameter of InfluxDBClient.

For more info about how configure HTTP retry see details in urllib3 documentation.

from urllib3 import Retry

from influxdb_client import InfluxDBClient

retries = Retry(connect=5, read=2, redirect=5)
client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org", retries=retries)

Nanosecond precision

The Python's datetime doesn't support precision with nanoseconds so the library during writes and queries ignores everything after microseconds.

If you would like to use datetime with nanosecond precision you should use pandas.Timestamp that is replacement for python datetime.datetime object, and also you should set a proper DateTimeHelper to the client.

from influxdb_client import Point, InfluxDBClient
from influxdb_client.client.util.date_utils_pandas import PandasDateTimeHelper
from influxdb_client.client.write_api import SYNCHRONOUS

"""
Set PandasDate helper which supports nanoseconds.
"""
import influxdb_client.client.util.date_utils as date_utils

date_utils.date_helper = PandasDateTimeHelper()

"""
Prepare client.
"""
client = InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org")

write_api = client.write_api(write_options=SYNCHRONOUS)
query_api = client.query_api()

"""
Prepare data
"""

point = Point("h2o_feet") \
    .field("water_level", 10) \
    .tag("location", "pacific") \
    .time('1996-02-25T21:20:00.001001231Z')

print(f'Time serialized with nanosecond precision: {point.to_line_protocol()}')
print()

write_api.write(bucket="my-bucket", record=point)

"""
Query: using Stream
"""
query = '''
from(bucket:"my-bucket")
        |> range(start: 0, stop: now())
        |> filter(fn: (r) => r._measurement == "h2o_feet")
'''
records = query_api.query_stream(query)

for record in records:
    print(f'Temperature in {record["location"]} is {record["_value"]} at time: {record["_time"]}')

"""
Close client
"""
client.close()

How to use Asyncio

Starting from version 1.27.0 for Python 3.7+ the influxdb-client package supports async/await based on asyncio, aiohttp and aiocsv. You can install aiohttp and aiocsv directly:

$ python -m pip install influxdb-client aiohttp aiocsv

or use the [async] extra:

$ python -m pip install influxdb-client[async]

:warning:

The InfluxDBClientAsync should be initialised inside async coroutine otherwise there can be unexpected behaviour. For more info see: Why is creating a ClientSession outside an event loop dangerous?.

Async APIs

All async APIs are available via influxdb_client.client.influxdb_client_async.InfluxDBClientAsync. The async version of the client supports following asynchronous APIs:

and also check to readiness of the InfluxDB via /ping endpoint:

The InfluxDBClientAsync constructor accepts a number of configuration properties. Most useful among these are:

import asyncio

from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async def main():
    async with InfluxDBClientAsync(url="http://localhost:8086", token="my-token", org="my-org") as client:
        ready = await client.ping()
        print(f"InfluxDB: {ready}")

if __name__ == "__main__":
    asyncio.run(main())

Async Write API

The influxdb_client.client.write_api_async.WriteApiAsync supports ingesting data as:

import asyncio

from influxdb_client import Point
from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async def main():
    async with InfluxDBClientAsync(url="http://localhost:8086", token="my-token", org="my-org") as client:

        write_api = client.write_api()

        _point1 = Point("async_m").tag("location", "Prague").field("temperature", 25.3)
        _point2 = Point("async_m").tag("location", "New York").field("temperature", 24.3)

        successfully = await write_api.write(bucket="my-bucket", record=[_point1, _point2])

        print(f" > successfully: {successfully}")

if __name__ == "__main__":
    asyncio.run(main())

Async Query API

The influxdb_client.client.query_api_async.QueryApiAsync supports retrieve data as:

import asyncio

from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async def main():
    async with InfluxDBClientAsync(url="http://localhost:8086", token="my-token", org="my-org") as client:
        # Stream of FluxRecords
        query_api = client.query_api()
        records = await query_api.query_stream('from(bucket:"my-bucket") '
                                               '|> range(start: -10m) '
                                               '|> filter(fn: (r) => r["_measurement"] == "async_m")')
        async for record in records:
            print(record)

if __name__ == "__main__":
    asyncio.run(main())

Async Delete API

import asyncio
from datetime import datetime

from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async def main():
    async with InfluxDBClientAsync(url="http://localhost:8086", token="my-token", org="my-org") as client:
        start = datetime.fromtimestamp(0)
        stop = datetime.now()
        # Delete data with location = 'Prague'
        successfully = await client.delete_api().delete(start=start, stop=stop, bucket="my-bucket",
                                                        predicate="location = \"Prague\"")
        print(f" > successfully: {successfully}")

if __name__ == "__main__":
    asyncio.run(main())

Management API

import asyncio

from influxdb_client import OrganizationsService
from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async def main():
    async with InfluxDBClientAsync(url='http://localhost:8086', token='my-token', org='my-org') as client:
        # Initialize async OrganizationsService
        organizations_service = OrganizationsService(api_client=client.api_client)

        # Find organization with name 'my-org'
        organizations = await organizations_service.get_orgs(org='my-org')
        for organization in organizations.orgs:
            print(f'name: {organization.name}, id: {organization.id}')

if __name__ == "__main__":
    asyncio.run(main())

Proxy and redirects

You can configure the client to tunnel requests through an HTTP proxy. The following proxy options are supported:

from influxdb_client.client.influxdb_client_async import InfluxDBClientAsync

async with InfluxDBClientAsync(url="http://localhost:8086",
                               token="my-token",
                               org="my-org",
                               proxy="http://localhost:3128") as client:

If your proxy notify the client with permanent redirect (HTTP 301) to different host. The client removes Authorization header, because otherwise the contents of Authorization is sent to third parties which is a security vulnerability.

Client automatically follows HTTP redirects. The default redirect policy is to follow up to 10 consecutive requests. The redirects can be configured via:

Logging

The client uses Python's logging facility for logging the library activity. The following logger categories are exposed:

The default logging level is warning without configured logger output. You can use the standard logger interface to change the log level and handler:

import logging
import sys

from influxdb_client import InfluxDBClient

with InfluxDBClient(url="http://localhost:8086", token="my-token", org="my-org") as client:
    for _, logger in client.conf.loggers.items():
        logger.setLevel(logging.DEBUG)
        logger.addHandler(logging.StreamHandler(sys.stdout))

Debugging

For debug purpose you can enable verbose logging of HTTP requests and set the debug level to all client's logger categories by:

client = InfluxDBClient(url="http://localhost:8086", token="my-token", debug=True)

Both HTTP request headers and body will be logged to standard output.

Local tests

# start/restart InfluxDB2 on local machine using docker
./scripts/influxdb-restart.sh

# install requirements
pip install -e . --user
pip install -e .\[extra\] --user
pip install -e .\[test\] --user

# run unit & integration tests
pytest tests

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/influxdata/influxdb-client-python.

License

The gem is available as open source under the terms of the MIT License.