influx-client

InfluxDB client compatible with 1.5. This client uses the awesome requests library to provide connection pooling for each unique InfluxDB URL given.

This InfluxDB client is created, maintained, and supported by Axiom Exergy.

Prerequisites

This client has only been tested and used against InfluxDB 1.5 and Python 3.5. If you want to support any other environments, please submit a pull request.

Installation

You can install this client via PyPI:

$ pip install influx-client

Or by cloning this repository:

$ git clone https://github.com/AxiomExergy/influx-client.git
$ cd influx-client
$ pip install .  # For a regular install
$ python setup.py develop  # OR for a development install

Usage

This section describes basic usage.

Quickstart Example

This InfluxDB client is designed to be very simple. It takes a URL to the InfluxDB API on creation, and otherwise supplies all parameters per write() call.

If the InfluxDB API returns an error that the chosen database does not exist, the client will issue a CREATE DATABASE ... query, followed by retrying the write request.

from influx import InfluxDB

# This creates the client instance... subsequent calls with the same URL will
# return the exact same instance, allowing you to use socket pooling for faster
# requests with less resources.
client = InfluxDB('http://127.0.0.1:8086')

# Creating the database is optional - calls to write() will try to create the
# database if it does not exist.
client.create_database('mydatabase')

# You can write as many fields and tags as you like, or override the *time* for
# the data points
client.write('mydatabase', 'mymeasurement', fields={'value': 1.0},
             tags={'env': 'example'})

# You can write multiple datapoints at a time
client.write_many('mydatabase', 'mymeasurement', fields=['value', 'alpha'],
                  values=[[1.0, 0.5], [1.1, 0.6]], tags={'env': 'example'})

# You can query for data relative to now()
data = client.select_recent('mydatabase', 'mymeasurement', time_relative='1h')

# You can query with arbitrary WHERE clauses and LIMITs
data = client.select_where('mydatabase', 'mymeasurement', where='time > 0', limit=1)

# You can clean up after yourself, for example in testing environments
client.drop_measurement('mymeasurement', 'mydatabase')

# You can also drop the entire database, if necessary
client.drop_database('mydatabase')

# Subsequent client creation will give the same instance
client2 = InfluxDB('http://127.0.0.1:8086')
client is client2  # This is True

Development

This section describes development and contribution for influx-client.

• Development is on GitHub.
• Installation is via PyPI.
• Issues are on Github.
• Releases are on GitHub.

Contributors

This section lists everyone who has contributed to this project.

• shakefu (Creator, Maintainer)

Repository Layout

There are a few important pieces in this repository:

• influx/ - The influx Python package
• test/ - Python nosetests
• Dockerfile, docker-compose.yml - Docker configuration for testing
• LICENSE, README.md - Documentation and legal

Running Tests

You can run the full test suite with supporting InfluxDB instance using docker-compose.

The following command will build the test image and run all tests:

docker-compose up --build --force-recreate --remove-orphans --exit-code-from influx

When tests are complete, you can clean up supporting services using:

docker-compose down

Making Pull Requests

Pull requests must pass CI to be considered for inclusion. If your pull request does not have squashed commits, your commits should follow the topic: description style. See the commit history for examples.

API

This section describes the public API for influx-client.

influx.client(url, timeout=60, precision='u')

Helper method to allow you to instantiate an InfluxDB client directly from the top level package.

• url (str) - URL to InfluxDB API (required)
• timeout (int, default 60) - Timeout in seconds for requests
• precision (str, default 'u') - Precision string to use for querying

InfluxDB(url, timeout=60, precision='u')

This is the main InfluxDB client. It works as a singleton instance per url. In threaded or event loop based environments it relies on the requests library connection pooling (which in turn relies on urllib3) for thread safety.

• url (str) - URL to InfluxDB API (such as 'http://127.0.0.1:8086')
• timeout (int, default 60) - Timeout in seconds for requests
• precision (str, default 'u') - Precision string to use for querying InfluxDB. ([See the documentation] (https://docs.influxdata.com/influxdb/v1.5/tools/api/#query) for what is available.)

.create_database(database)

Issues a CREATE DATABASE ... request to the InfluxDB API. This is an idempotent operation.

• database (str) - Database name

.drop_database(database)

Issues a DROP DATABASE ... request to the InfluxDB API. This will raise a 404 HTTPError if the database does not exist.

• database (str) - Database name

.drop_measurement(measurement, database)

Issues a DROP MEASUREMENT ... request to the InfluxDB API for the specified database.

• measurement (str) - Measurement name
• database (str) - Database in which measurement resides

.write(database, measurement, fields, tags={}, time=None)

Write data points to the specified database and measurement.

• database (str) - Database name
• measurement (str) - Measurement name
• fields (dict) - Dictionary of field_name: value data points
• tags (dict, optional) - Dictionary of tag_name: value tags to associate with the data points
• time (datetime, optional) - Datetime to use instead of InfluxDB's server-side "now"

.write_many(database, measurement, fields, values, tags={}, time_field=None)

Write data points to the specified database and measurement.

• database (str) - Database name
• measurement (str) - Measurement name
• fields (list) - List of field names, ordered the same as values
• values (list) - List of values (list of lists)
• tags (dict, optional) - Dictionary of tag_name: value tags to associate with the data points
• time_field (str, optional) - Field name to extract and use as timestamp

.select_recent(database, measurement, fields='*', tags={}, relative_time='15m')

Query the InfluxDB API for measurement in database, using the fields string, limited to matching tags for the recent relative_time.

Returns the raw JSON response from InfluxDB.

• database (str) - Database name
• measurement (str) - Measurement name
• fields (str, default '*') - String formatted fields for SELECT query
• tags (dict, optional) - Dictionary of tag_name: value tags to match
• relative_time (str, default '15m') - Relative time string

.select_where(database, measurement, fields='*', tags={}, where='time > now() - 15m', desc=False, limit=None)

Query the InfluxDB API for measurement in database, using the fields string, limited to matching tags with the where clause and limit applied.

Returns the raw JSON response from InfluxDB.

• database (str) - Database name
• measurement (str) - Measurement name
• fields (str, default '*') - String formatted fields for SELECT query
• tags (dict, optional) - Dictionary of tag_name: value tags to match
• where (str, default 'time > now() - 15m') Where clause to add
• desc (bool, default False) Add the ORDER BY time DESC clause
• limit (int, optional) Limit to this number of data points

.select_into([database,] target, source, fields='*', where=None, group_by='*')

Returns count of data points moved by a SELECT ... INTO ... FROM ... query.

The query will follow the format:

SELECT *fields* INTO *target* FROM *source* WHERE *where* GROUP BY *group_by*

The WHERE and GROUP BY clauses are optional.

Note that if you do not include GROUP BY * or explicitly call out tags, tags will be recorded as fields.

• database (str) Database name (optional)
• target (str) Target measurement
• source (str) Source measurement
• fields (str) Fields portion of the SELECT clause (optional, default: '*')
• where (str) WHERE portion of the SELECT clause (optional)
• group_by (str) GROUP BY portion of the SELECT clause (optional, default: '*')

This method may be called with variable arguments, if you wish to specify the full qualified measurement name with database.

Example:

InfluxDB().select_into('database', 'target', 'source')
# Is the same as
InfluxDB().select_into('database.default.target', 'database.default.source')
# Where often the default is 'autogen'

.show_tags(database, measurement)

Query the InfluxDB API and return a list of tag names in database and measurement.

• database (str) - Database name
• measurement (str) - Measurement name

.show_fields(database, measurement)

Query the InfluxDB API and return a list of field names in database and measurement.

• database (str) - Database name
• measurement (str) - Measurement name

License

This repository and its codebase are made public under the Apache License v2.0. We ask that if you do use this work please attribute Axiom Exergy and link to the original repository.

Changelog

See Releases for detailed release notes.