Usage

CLI

Authentication

The CloudKitty client can either be used through the standalone CLI executable (cloudkitty) or through the OpenStack Client module (openstack rating).

When using CloudKitty in standalone mode (ie without Keystone authentication), the API endpoint and the auth method must be specified:

cloudkitty --os-endpoint http://cloudkitty-api:8889 --os-auth-type cloudkitty-noauth module list

These options can also be specified as environment variables:

export OS_ENDPOINT=http://cloudkitty-api:8889
export OS_AUTH_TYPE=cloudkitty-noauth
cloudkitty module list

The exact same options apply when using the OpenStack Client plugin:

# EITHER
openstack rating --os-endpoint http://cloudkitty-api:8889 --os-auth-type cloudkitty-noauth module list

# OR
export OS_ENDPOINT=http://cloudkitty-api:8889
export OS_AUTH_TYPE=cloudkitty-noauth
openstack rating module list

Version

Two versions of the client exist: v1 and v2. The v2 version adds support for v2 API endpoints. The default API version is 1. You can specify which API version you want to use via a CLI option:

# EITHER
cloudkitty --os-rating-api-version 2 summary get

# OR
export OS_RATING_API_VERSION=2
cloudkitty summary get

Again, the option can also be provided to the OSC plugin, both via the CLI flag or the environment variable.

Python library

You can use cloudkittyclient with or without keystone authentication. In order to use it without keystone authentication, cloudkittyclient provides the CloudKittyNoAuthPlugin keystoneauth plugin:

>>> from cloudkittyclient import client as ck_client
>>> from cloudkittyclient import auth as ck_auth

>>> auth = ck_auth.CloudKittyNoAuthPlugin(endpoint='http://127.0.0.1:8889')
>>> client = ck_client.Client('1', auth=auth)
>>> client.report.get_summary()
{u'summary': [{u'begin': u'2018-03-01T00:00:00',
               u'end': u'2018-04-01T00:00:00',
               u'rate': u'1672.71269',
               u'res_type': u'ALL',
               u'tenant_id': u'bea6a24f77e946b0a92dca7c78b7870b'}]}

Else, use it the same way as any other OpenStack client:

>>> import os

>>> from keystoneauth1 import session
>>> from keystoneauth1.identity import v3

>>> from cloudkittyclient import client as ck_client

>>> auth = v3.Password(
       auth_url=os.environ.get('OS_AUTH_URL'),
       project_domain_id=os.environ.get('OS_PROJECT_DOMAIN_ID'),
       user_domain_id=os.environ.get('OS_USER_DOMAIN_ID'),
       username=os.environ.get('OS_USERNAME'),
       project_name=os.environ.get('OS_PROJECT_NAME'),
       password=os.environ.get('OS_PASSWORD'))

>>> ck_session = session.Session(auth=auth)

>>> c = ck_client.Client('1', session=ck_session)

>>> c.report.get_summary()
{u'summary': [{u'begin': u'2018-03-01T00:00:00',
                u'end': u'2018-04-01T00:00:00',
                u'rate': u'1672.71269',
                u'res_type': u'ALL',
                u'tenant_id': u'bea6a24f77e946b0a92dca7c78b7870b'}]}

Warning

If you want to use SSL with the client as a python library, you need to provide a cert to keystone’s session object. Else, two additional options are available if you provide an auth object to the client: insecure and cacert:

>>> client = ck_client.Client(
        '1', auth=auth, insecure=False, cacert='/path/to/ca')

If you want to use the v2 API, you have to specify it at client instanciation

c = ck_client.Client('2', session=session)

When using the cloudkitty CLI client with keystone authentication, the auth plugin to use should automagically be detected. If not, you can specify the auth plugin to use with --os-auth-type/--os-auth-plugin:

$ cloudkitty --debug --os-auth-type cloudkitty-noauth summary get
+------------+---------------+------------+---------------------+---------------------+
| Project ID | Resource Type | Rate       | Begin Time          | End Time            |
+------------+---------------+------------+---------------------+---------------------+
| ALL        | ALL           | 1676.95499 | 2018-03-01T00:00:00 | 2018-04-01T00:00:00 |
+------------+---------------+------------+---------------------+---------------------+

CSV report generation

An output formatter (DataframeToCsvFormatter) has been created in order to allow CSV report generation through the client. It can be used with the -f df-to-csv option.

$ cloudkitty dataframes get -b 2018-03-22T12:00:00 -f df-to-csv
Begin,End,Metric Type,Qty,Cost,Project ID,Resource ID,User ID
2018-03-01T12:00:00,2018-03-01T13:00:00,compute,1,2.0,53c3fe396a1a4ab0914b9aa997a5ff88,382d23c3-7b77-4e32-8d65-b3baf86ed7bb,38c1949c2e624f729b30e034ac787640
[...]

Warning

The df-to-csv formatter should NEVER be used together with the -c/--column option and should only be used for the dataframes get command.

The example above shows how to get a CSV report with the standard columns. If you want other columns, it is possible to customize the formatter through a configuration file:

# This exact file format must be respected (- column_name: json_path)
# The path is interpreted using jsonpath-rw-ext, see
# https://github.com/sileht/python-jsonpath-rw-ext for syntax reference
- 'Begin': '$.begin'
- 'End': '$.end'
- 'User ID': '$.desc.user_id'
- 'Resource ID': '$.desc.resource_id'
- 'Qty': '$.volume'
- 'Cost': '$.rating'

Example with this config file:

$ cloudkitty dataframes get -f df-to-csv --format-config-file /etc/cloudkitty/csv_config.yml > report.csv
$ head -n 2 report.csv
Begin,End,User ID,Resource ID,Qty,Cost
2018-03-01T12:00:00,2018-03-01T13:00:00,38c1949c2e624f729b30e034ac787640,382d23c3-7b77-4e32-8d65-b3baf86ed7bb,1,2.0

An other config file is provided: legacy_csv_config.yml. This file is compatible with the format of cloudkitty-writer’s CSV reports.