Dive in! …to get started using the dnacentersdk package:

Make sure that you have:

Get your DNA Center Access Token

To interact with the DNA Center APIs, you must have a DNA Center Access Token. A DNA Center Access Token is how the DNA Center APIs validate access and identify the requesting user.

As a best practice, you can store your DNA Center ‘credentials’ as an environment variables in your development or production environment.

By default, dnacentersdk will look for the following environment variables to create new connection objects:

  • DNA_CENTER_DEBUG - Tells the SDK whether to log request and response information. Useful for debugging and seeing what is going on under the hood. Defaults to False.

  • DNA_CENTER_VERSION - DNA Center API version to use. Defaults to ‘’.

  • DNA_CENTER_ENCODED_AUTH - It takes priority. It is the username:password encoded in base 64. For example ‘ZGV2bmV0dXNlcjpDaXNjbzEyMyEK’ which decoded is ‘devnetuser:Cisco123!’

  • DNA_CENTER_USERNAME - HTTP Basic Auth username.

  • DNA_CENTER_PASSWORD - HTTP Basic Auth password.

  • DNA_CENTER_BASE_URL - The base URL to be prefixed to the individual API endpoint suffixes. Defaults to ‘https://sandboxdnac.cisco.com:443’.

  • DNA_CENTER_SINGLE_REQUEST_TIMEOUT - Timeout (in seconds) for RESTful HTTP requests. Defaults to 60.

  • DNA_CENTER_WAIT_ON_RATE_LIMIT - Enables or disables automatic rate-limit handling. Defaults to True.

  • DNA_CENTER_VERIFY - Controls whether to verify the server’s TLS certificate or not. Defaults to True.

However, you choose to set it, if you have DNA_CENTER_VERSION, DNA_CENTER_USERNAME and DNA_CENTER_PASSWORD, or DNA_CENTER_VERSION and DNA_CENTER_ENCODED_AUTH environment variables, you are good to go. dnacentersdk will use them to create your access token when creating new DNACenterAPI objects.

If you don’t want to set your credentials as environment variables, you can manually provide them as parameters when creating a DNACenterAPI object.

Set credentials as environment variables

There are many places and diverse ways that you can set an environment variable, which can include:

  • A setting within your development IDE

  • A setting in your container / PaaS service

  • A statement in a shell script that configures and launches your app

It can be as simple as setting it in your CLI before running your script…

$ DNA_CENTER_USERNAME=your_username_here
$ DNA_CENTER_PASSWORD=your_password_here
$ python myscript.py

…or putting your credentials in a shell script that you source when your shell starts up or before your run a script:

$ cat mycredentials.sh
export DNA_CENTER_ENCODED_AUTH=your_encoded_auth_here
$ source mycredentials.sh
$ python myscript.py

Create a DNACenterAPI “Connection Object”

To make interacting with the DNA Center APIs as simple and intuitive as possible, all of the APIs have ‘wrapped’ underneath a single interface. To get started, import the DNACenterAPI class and create an API “connection object”.

>>> from dnacentersdk import DNACenterAPI
>>> api = DNACenterAPI()

As discussed above (Get your DNA Center Access Token), dnacentersdk defaults to pulling from environment variables to generate your access token. If you do not have those environment variables set and you try to create a new DNACenterAPI object without providing them, a AccessTokenError will be raised (a dnacentersdkException subclass).

>>> from dnacentersdk import DNACenterAPI
>>> api = DNACenterAPI()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "dnacentersdk/__init__.py", line 237, in __init__
    raise AccessTokenError(error_message)
AccessTokenError: You need an access token to interact with the DNA Center
APIs. DNA Center uses HTTP Basic Auth to create an access
token. You must provide the username and password or just
the encoded_auth, either by setting each parameter or its
environment variable counterpart (DNA_CENTER_USERNAME,

If you don’t provide a known version and try to create a new DNACenterAPI, a VersionError will be raised.

>>> from dnacentersdk import DNACenterAPI
>>> api = DNACenterAPI(username='devnetuser', password='Cisco123!', base_url='https://sandboxdnac.cisco.com:443', version='0.1.12')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "dnacentersdk/__init__.py", line 209, in __init__
    raise VersionError(error_message)
VersionError: Unknown API version, known versions are,,, and

Use the arguments to manually provide enough information for the HTTP Basic Auth process, when creating a new DNACenterAPI connection object.

>>> from dnacentersdk import DNACenterAPI
>>> # Create a DNACenterAPI connection object; it uses DNA Center sandbox URL and encoded_auth, with DNA Center API version
>>> api = DNACenterAPI(encoded_auth='ZGV2bmV0dXNlcjpDaXNjbzEyMyEK', base_url="https://sandboxdnac.cisco.com:443", version='')
>>> from dnacentersdk import DNACenterAPI
>>> # Create a DNACenterAPI connection object; it uses DNA Center username and password, with DNA Center API version
>>> # The base_url used by default is `from dnacentersdk.config import DEFAULT_BASE_URL`
>>> api = DNACenterAPI(username='devnetuser', password='Cisco123!', base_url="https://sandboxdnac.cisco.com:443", version='')

Note that this can be very useful if you are reading authentication credentials from a file or database and/or when you want to create more than one connection object.

>>> from dnacentersdk import DNACenterAPI
>>> kingston_auth = 'ZG5hY2VudGVydXNlcjpDaXNjbzEyMyEK'
>>> london_auth = ('london', 'rcx0cf43!')
>>> kingston_api = DNACenterAPI(encoded_auth=kingston_auth, base_url="https://sandboxdnac.cisco.com:443", version='')
>>> london_api = DNACenterAPI(*london_auth, base_url="", version='')  # * Unpacks tuple


Besides username, password, encoded_auth, base_url, and version, there are other parameters when creating the DNACenterAPI, many of them have a default value (check Package Constants for more).

When dealing with certificates, the most important one is the verify parameter.

To avoid getting errors like the following:

>>> from dnacentersdk import DNACenterAPI
>>> own_api = DNACenterAPI(encoded_auth='dXNlcm5hbWU6cGFzc3dvcmQK',
... base_url="", version='')
requests.exceptions.SLError: HTTPSConnectionPool(host='', port=443):
Max retries exceeded with url: /dna/system/api/v1/auth/token (Caused by
SSLError (SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate
verify failed: self signed certificate in certificate chain (_ssl.c:1076)')))

Include the verify parameter and set it to False:

>>> from dnacentersdk import DNACenterAPI
>>> own_api = DNACenterAPI(encoded_auth='dXNlcm5hbWU6cGFzc3dvcmQK',
... base_url="", version='1.3.0',
... verify=False)
InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate
 verification is strongly advised. See: https://urllib3.readthedocs.io/en/latest/advanced-usage.html#ssl-warnings

You will see urllib3 warnings instead. If you want to disable them, the easiest way is with:

>>> import urllib3
>>> urllib3.disable_warnings()

Package Constants

The following are the default values pulled from dnacentersdk.config and used when creating the connection object.


debug default value.


version default value.

DEFAULT_BASE_URL = 'https://sandboxdnac.cisco.com:443'

base_url default value.


single_request_timeout default value. Timeout (in seconds) for the RESTful HTTP requests.


wait_on_rate_limit default value. Enables or disables automatic rate-limit handling.


verify default value. Controls whether to verify the server’s TLS certificate or not.

Making API Calls

Now that you have created a DNACenterAPI “connection object,” you are ready to start making API calls.

>>> api.device_onboarding_pnp.get_workflows()[0] # api.pnp.get_workflows()[0] till version 1.3.0
  'version': 1,
  'name': 'test',
  'description': '',
  'useState': 'Available',
  'type': 'Standard',
  'addedOn': 1559538796969,
  'lastupdateOn': 1560455244769,
  'startTime': 0,
  'endTime': 0,
  'execTime': 0,
  'currTaskIdx': 0,
  'tasks': [{'taskSeqNo': 0,
    'name': 'Config Download',
    'type': 'Config',
    'startTime': 0,
    'endTime': 0,
    'timeTaken': 0,
    'currWorkItemIdx': 0,
    'workItemList': [],
    'configInfo': {'configId': '62066037-55cd-4c05-8e09-45674994a274',
    'configFileUrl': None,
    'fileServiceId': None,
    'saveToStartUp': True,
    'connLossRollBack': True,
    'configParameters': None}}],
  'addToInventory': True,
  'tenantId': '5bd3634ab2bea0004c3ebb58',
  'id': '5cf4ac6c568ecc000779da5c'

It really is that easy.

All of the calls have been wrapped and represented as native Python method calls, like DNACenterAPI.device_onboarding_pnp.get_workflows() which gets the workflows details for the pnp - see the Get Workflows API endpoint documentation.

As you can see, we have represented the API endpoints using simple terms that are aligned with the API docs; for example, representing the /onboarding/pnp-workflow API endpoint as a device_onboarding_pnp.get_workflows() method available underneath the DNACenterAPI connection object.

A full list of the available API methods, with their descriptions and parameters, is available in the User API Doc.

A summary of the structure is available for each version supported

You can easily access and call any of these methods directly from your DNACenterAPI connection object:

>>> api.device_onboarding_pnp.get_device_list(limit=1)  # api.pnp.get_device_list(limit=1) till version 1.3.0
[{'version': 1,
  'deviceInfo': {'serialNumber': '1234567890s',
  'name': 'Postname-add',
  'pid': 'ws-c9300',
  'lastSyncTime': 0,
  'addedOn': 1559870763581,
  'lastUpdateOn': 1559870763581,
  'firstContact': 0,
  'lastContact': 0,
  'state': 'Unclaimed',
  'onbState': 'Not Contacted',
  'cmState': 'Not Contacted',
  'source': 'User',
  'reloadRequested': False,
  'aaaCredentials': {'username': '', 'password': ''},
  'populateInventory': False,
  'stack': False,
  'sudiRequired': False,
  'validActions': {'editSUDI': True,
    'editWfParams': True,
    'delete': True,
    'claim': True,
    'unclaim': True,
    'reset': False}},
  'workflowParameters': {},
  'runSummaryList': [{'timestamp': 1559870763581,
    'details': 'User Added Device',
    'errorFlag': False}],
  'tenantId': '5bd3634ab2bea0004c3ebb58',
  'id': '5cf9bd2b568ecc000779da65'}]

Catching Exceptions

If something should go wrong with the API call, an exception will be raised. ApiError exceptions are raised when an error condition is returned from the DNA Center cloud. Details will be provided in the error message.

>>> from dnacentersdk import DNACenterAPI, ApiError
>>> api = DNACenterAPI(username='devnetuser', password='Cisco123!')
>>> # The base_url used by default is `from dnacentersdk.config import DEFAULT_BASE_URL`
>>> task = api.task.get_task_by_operationid(
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "dnacentersdk/api/task.py", line 584, in get_task_by_operationid
  File "dnacentersdk/restsession.py", line 280, in get
    response = self.request('GET', url, erc, params=params, **kwargs)
  File "dnacentersdk/restsession.py", line 236, in request
    check_response_code(response, erc)
  File "dnacentersdk/utils.py", line 217, in check_response_code
    raise ApiError(response)
dnacentersdk.exceptions.ApiError: [500] Server Error - errorId=20,
componentName=CRUD executeQuery Failed! errorId=20,componentName=CRUD
executeQuery Failed! Named query not known: task.findTaskByOperationId

You can catch any errors returned by the DNA Center cloud by catching ApiError exceptions in a try-except block.

>>> from dnacentersdk.exceptions import ApiError
>>> try:
...     task = api.task.get_task_by_operationid(
...       limit=2,
...       offset=1,
...       operation_id='xyz')
... except ApiError as e:
...     print(e)
ApiError: [500] Server Error - errorId=20,componentName=CRUD executeQuery Failed! errorId=20,
componentName=CRUD executeQuery Failed! Named query not known: task.findTaskByOperationId

dnacentersdk will also raise a number of other standard errors (TypeError, ValueError, etc.); however, these errors are usually caused by incorrect use of the package or methods and should be sorted while debugging your app.

Working with Returned Objects

The DNA Center cloud returns data objects in JSON format, like so:

[{ "version": 1,
  "deviceInfo": { "serialNumber":  "1234567890s",
  "name":  "Postname-add ",
  "pid":  "ws-c9300 ",
  "lastSyncTime": 0,
  "addedOn": 1559870763581,
  "lastUpdateOn": 1559870763581,
  "firstContact": 0,
  "lastContact": 0,
  "state":  "Unclaimed ",
  "onbState":  "Not Contacted ",
  "cmState":  "Not Contacted ",
  "source":  "User ",
  "reloadRequested": false,
  "aaaCredentials": { "username":  "",  "password":  ""},
  "populateInventory": false,
  "stack": false,
  "sudiRequired": false,
  "validActions": { "editSUDI": true,
    "editWfParams": true,
    "delete": true,
    "claim": true,
    "unclaim": true,
    "reset": false}},
  "workflowParameters": {},
  "runSummaryList": [{ "timestamp": 1559870763581,
    "details":  "User Added Device ",
    "errorFlag": false}],
  "tenantId":  "5bd3634ab2bea0004c3ebb58 ",
  "id":  "5cf9bd2b568ecc000779da65 "}]

Sure, JSON data objects can easily be parsed and represented in Python using dictionaries, but when working with an ‘object’ wouldn’t it be nice to be able to work with it like an object - using native object syntax (like accessing attributes using ‘.’ notation)? dnacentersdk enables you to do just that:

>>> pnp_devices = api.device_onboarding_pnp.get_device_list() # api.device_onboarding_pnp.get_device_list() till version 1.3.0
>>> pnp_devices[0].id
>>> pnp_devices[0].deviceInfo.state
>>> pnp_devices[0].deviceInfo.serialNumber

Representing and treating DNA Center data objects as Python data objects, can really help clean up your code and make coding easier:

  1. You don’t need to create variables to hold the data attributes, just use the attributes available underneath the data object.

    >>> # Do this
    >>> api.device_onboarding_pnp.get_device_history(serial_number=pnp_devices[0].deviceInfo.serialNumber)
    {'response': [{'timestamp': 1559870763581, 'details': 'User Added Device', 'errorFlag': False}], 'statusCode': 200}
    >>> # Instead of this
    >>> device_serialNumber = pnp_devices[0].deviceInfo.serialNumber
    >>> api.device_onboarding_pnp.get_device_history(serial_number=device_serialNumber)
    {'response': [{'timestamp': 1559870763581, 'details': 'User Added Device', 'errorFlag': False}], 'statusCode': 200}
  2. When accessing ‘optional’ attributes, like pnp_devices[0].workflowParameters.configList attribute of DNA Center PnP object, the response object will return None when the attribute is not present and will return the attribute’s value when it is present. This avoids some boiler plate code and/or needless exception handling, when working with optional attributes.

    >>> # Instead of doing this
    >>> for d in devices:
    ...     if hasattr(d, 'workflowParameters') and hasattr(d.workflowParameters, 'configList'):
    ...         # Do something with the configList attribute
    ...         pass
    >>> # Or this
    >>> try:
    ...     for d in devices:
    ...         # Do something with the configList attribute
    ...         d.workflowParameters.configList
    ... except AttributeError as e:
    ...     pass
    >>> # You can do this, which is cleaner
    >>> for d in devices:
    ...     if d.workflowParameters and d.workflowParameters.configList:
    ...         # Do something with the configList attribute
    ...         pass
  3. It just feels more natural. :-) When iterating through sequences, and working with objects in those sequences (see the next section), working with objects as objects is definitely more Pythonic.

    The Zen of Python (PEP 20):

    “Beautiful is better than ugly.” “Simple is better than complex.”

The currently modeled DNA Center Data Object with its functions, is available here in the User API Doc.

What if DNA Center adds new data attributes?

Attribute access WILL WORK for the newly added attributes (yes, without a package update!). dnacentersdk is written to automatically take advantage of new attributes and data as they are returned.

Configuring Logging for dnacentersdk

The main dnacentersdk logger is dnacentersdk.

Other loggers are dnacentersdk.exceptions, dnacentersdk.restsession and dnacentersdk.api.custom_caller.

The dnacentersdk adds only the logging.NullHandler following the logging recommendations for libraries

So you can add your logging handlers according to your needs.

import logging
import warnings
from dnacentersdk import DNACenterAPI

# Another way to disable warnings caused by (verify=False)
warnings.filterwarnings('ignore', message='Unverified HTTPS request')

logger = logging.getLogger('simple_example')
# create console handler and set level to debug
ch = logging.StreamHandler()
# create formatter
formatter = logging.Formatter('%(asctime)s - %(name)s - %(levelname)s - %(message)s')
# add formatter to ch
# add ch to logger

ch_ = logging.StreamHandler()
api = DNACenterAPI(verify=False, debug=True)

logger.debug('simple message')

Adding API call definitions

Custom caller functions help you:

  1. Add support for custom API calls.

  2. Add support for API calls that are/were not documented when the SDK was released.

from dnacentersdk import api

# Create a DNACenterAPI connection object;
# it uses DNA Center sandbox URL, username and password, with DNA Center API version,
# and requests to verify the server's TLS certificate with verify=True.
api_ = api.DNACenterAPI(username="devnetuser",

# Add your custom API call to the connection object.
# Define the get_global_credentials function.
# Call it with:
#     get_global_credentials('NETCONF')
def get_global_credentials(subtype):
    return api_.custom_caller.call_api('GET',
                                   'credentialSubType': subtype

# Add your custom API call to the connection object.
# Define the delete_global_credentials_by_id function
# under the custom_caller wrapper.
# Call it with:
#     api_.custom_caller.delete_global_credentials_by_id('be456g16-14fd-4cac-94b7-ac3b8f9f')
                          lambda global_credential_id:
                                      'credentialId': global_credential_id,

# Advance usage example using Custom Caller functions.
def setup_custom():
    Defines the get_global_credentials and create_netconf_credentials functions
    under the custom_caller wrapper, and with help documentation
    in two different ways.

    Check that they have been added with
        'get_global_credentials' in dir(api_.custom_caller)
        'create_netconf_credentials' in dir(api_.custom_caller)

    Quickly check that you indeed have them as functions with
        type(getattr(api_.custom_caller, 'create_netconf_credentials'))
        type(getattr(api_.custom_caller, 'create_netconf_credentials'))

    Check the documentation with


    # Alternative 1: Definition with helper function.
    def _get_global_credentials(credential_type):
        """Custom global credential API call, returns response attribute

                credential_type(str): Credential type as CLI
                    / SNMPV2_READ_COMMUNITY /
                    SNMPV2_WRITE_COMMUNITY / SNMPV3 /
                    HTTP_WRITE / HTTP_READ / NETCONF.

                MyDict: JSON response. Access the object's properties by using
                the dot notation or the bracket notation.
        return api_.custom_caller.call_api(
                                        'credentialSubType': credential_type
    # Finally add the function as an attribute.
    api_.custom_caller.add_api('get_global_credentials', _get_global_credentials)

    # Alternative 2: Definition with lambda function.
                            lambda port:
                                        "netconfPort": port
    # Finally add the documentation
    api_.custom_caller.create_netconf_credentials.__doc__ = """
        Custom global credential API call to add NETCONF credentials

            port(string): Netconf port number

        Returns: JSON response.

Check out the Custom Caller documentation to begin using it.

Copyright (c) 2019-2021 Cisco Systems.