This Python based client for NSX for vSphere 6.x gets its API structure information (e.g. URLs, parameters, schema, etc.) from a RAML file which describes the NSX for vSphere REST API. It has been developed and tested for use with VMware NSX for vSphere 6.x.
The latest version of the NSX for vSphere 6.x RAML file can be found at http://github.com/vmware/nsxraml
Please read the bellow Version information. The 2.0 Version of
nsxramlclient is needed to support the new format of the nsxraml spec on
http://github.com/vmware/nsxraml that introduced a breaking change in
the way schemas are handled. If you are using the 1.x version of
nsxramlclient you will need to use the 6.1.4, 6.1.6 or 6.2.2 versions of
the nsx raml spec. In the 2.0 version the method
extract_resource_body_schema was replaced with
extract_resource_body_example
Also, in the NSXv 6.3 version of the nsx raml spec XML Comments and pre-filled content was introduced in the RAML spec to improve readability of the created HTML and PDF artifacts. This change may break existing code, so version 2.0.6 of the nsxramlclient now removes any XML comments and pre-filled content.
NSX-v 6.4 introduced a new response code 202 for the controller creation API (Issue [#22]). This version of nsxramlclient accepts 202 as a successful response code
This version of the nsxramlclient will now delete any XML comments found in the body content examples in the nsxraml spec. Previously nsxramlclient ran into a traceback if XML comments were used. Also, in this version nsxramlclient by default will remove all pre-filled content in the body content examples. This is needed because of a change in the nsxraml spec where the focus is now more shifting towards documentation, and pre-filled content helps in terms of readability in the examples. This behaviour can be controlled by using the remove_content= parameter in the extract_resource_body_example method.
Added fail_mode='' option in NsxClient to raise an exception instead of sys.exit when setting fail_mode='raise', or continue without an exception if set to fail_mode='continue'. Default is still fail_mode='exit' to preserve backwards compatibility. Read the changed 'Create a session object' section for mode details
Change in the dependencies to include 'oyopenssl' to make nsxramlclient easier to install on Windows
Change in the lxml dependency. We are now mandating lxml 3.6.0 or lower because of installation issues seen on Windows with lxml 3.6.1.
This version of nsxramlclient added support for repeating keys in the XML body without nested structure bellow. Before v2.0.1 nsxramlclient only supports repeating keys in the XML body if these repeating key themself hold multiple keys bellow them (nesting). This patch adds the support for repeating keys that only hold a values bellow them and no nested structure. This was needed to support the API call to create secondary IP Addresses on ESG router interfaces
This version of nsxramlclient is needed to support the new format of the
nsxraml spec on http://github.com/vmware/nsxraml that introduced a
breaking change in the way schemas are handled. In the new NSX-v RAML
spec schemas are now real schemas that can be used to check the
correctness of your XML document. In the earlier versions the schema was
used to return an XML example as a python Dict. The new way of
retrieving the XML example dict is by using the new method introduced in
the 2.0 version named extract_resource_body_example
This release introduces new helper methods:
read_all_pages: This Method reads all pages from the API Get for the
['virtualWires', 'pagedEdgeList'] displayNames. This e.g. helps to
collect all logical switches present in the system without having to
know the needed page file size
normalize_list_return: There are API calls in NSX-v were you are
getting a None object if no Object is present, a Dict type when only one
Object is present, and a List of Dicts when more than one Object is
present. E.g. for the retrieval of logical switches. When passing the
API return to this function it will normalize the return to be a list.
If the input is a None Object (no Object exists in NSX-v) an empty list
is return. When one Object is found, a List with the one Dict is
returned, when the input is a list, it is returned back unmodified
Initial versions
The following install instructions are based on Ubuntu 14.04 LTS, but installations on other Linux distributions or on the Apple MacOS should be relatively similar and should be familiar to those familiar with Python.
Check whether pip is installed
pip --versionIf pip is not installed, install it with apt-get or the package manager appropriate for your operating system.
sudo apt-get update
sudo apt-get -y install python-pipNow you can use pip to install the nsx raml client
sudo pip install nsxramlclientIn some cases the installation may fail because of missing dependencies. You may see the following message and will have to install the required packages
ERROR: /bin/sh: 1: xslt-config: not found ** make sure the development packages of libxml2 and libxslt are installed **
This example shows installing the dependencies using the apt package manager and the apt-get command. Once dependencies are installed you can retry the pip installation of the nsxramlclient shown above.
sudo apt-get install build-essential libssl-dev libffi-dev libxml2-dev libxslt-dev python-dev zlib1g-devAlso sometimes you might run into the issue of missing Python OpenSSL on your Ubuntu machine. In this case you will see the following error message when importing nsxramlclient.client:
ImportError: No module named OpenSSL.SSLif you encounter this, simply install the package python-openssl manually:
sudo apt-get install python-opensslIt is required to create a session object with which you will interact with the NSX REST API. This session object will then expose the create, read, update and delete (CRUD) methods of each NSX object as well as some helper methods that will be useful.
from nsxramlclient.client import NsxClient
nsxraml_file = '/raml/nsxraml/nsxvapiv614.raml'
nsxmanager = 'nsxmanager.invalid.org'
nsx_username = 'admin'
nsx_password = 'vmware'
client_session = NsxClient(nsxraml_file, nsxmanager, nsx_username,
nsx_password, debug=False, fail_mode='raise')The NsxClient class has the following initialization parameters:
"""
:param raml_file:
This mandatory parameter is the RAML file used as the basis of all URL
compositions. It allows the client to extract the body schema and convert the schema into python dictionaries.
:param nsxmanager:
This mandatory parameter is either the hostname or IP Address of the NSX Manager appliance.
:param nsx_username:
This mandatory parameter is the username on NSX Manager used for authentication to the NSX REST API running on the NSX Manager.
:param nsx_password:
This mandatory parameter is the password of the user used for authentication to the NSX REST API running on the NSX Manager.
:param debug: Optional:
If set to True, the client will print extensive HTTP session information to stdout.
Default: False
:param verify: Optional:
If set to True, the client will strictly verify the certificate passed by NSX Manager. It is recommmended in all production environments to use signed certificates for the NSX REST API. Please refer to the NSX for vSphere documentation for information on how to convert from the self-signed certificate to a signed certificate.
Default: False
:param suppress_warnings: Optional:
If set to True, the client will print out a warning if NSX Manager uses a self signed certificate.
Default: True
:param fail_mode: Optional:
If not set, the client will exit using sys.exit when receiving any error status code from NSX like 400.
If fail_mode is set to 'raise', the exception nsxramlclient.exceptions.NsxError will be raised with status
being the HTTP status code received and msg being the error message returned by NSX in the body. If set to
'continue', no error will be raised, and the status and body is returned like in successful cases.
Default: 'exit'
:return: Returns a NsxClient Session Object
"""After you initialized a session object you have access to the following methods:
- create: Sends a HTTP POST to NSX Manager. More details will follow later in this readme.
- read: Sends a HTTP GET to NSX Manager
- update: Sends a HTTP PUT to NSX Manager
- delete: Sends a HTTP DELETE to NSX Manager
- view_response: Each of the above methods returns a Python OrderedDictionary with the HTTP Status code, location header, NSX Object Id, eTag Header and Body. This method outputs the OrderedDict in human readable text to stdout.
- extract_resource_body_schema: DEPRECATION WARING: Use the method extract_resource_body_schema. In future version this will be removed This method will retrieve the body schema from the RAML File (if the method has a body schema like most create methods), and will return a template python dictionary that can be used to construct subsequent API calls.
- extract_resource_body_example: This method will retrieve the body example from the RAML File (if the method has a body example like most create methods), and will return a template python dictionary that can be used to construct subsequent API calls.
- view_resource_body_schema: This method retrieves the body schema from the RAML file and outputs it to stdout as a pretty printed XML document.
- view_resource_body_example: This method retrieves the body example from the RAML file and outputs it to stdout as a pretty printed XML document.
- view_body_dict: This method takes a body dictionary (any python dictionary), and outputs it in a human readable format to stdout.
- view_resource_display_names: This method outputs displayNames and descriptions of all resources in the RAML File with their associated URI & query parameters, additional headers, and what methods are supported.
In [1]: client_session.read('vCenterStatus')
Out[2]: OrderedDict([('status', 200), ('body', {'vcConfigStatus': {'connected': 'true', 'lastInventorySyncTime': '1440444721014'}}), ('location', None), ('objectId', None), ('Etag', None)])The create, read, update and delete methods return a Python OrderedDict
with the following key/value pairs: - status: The HTTP status code
returned as an integer. - body: The response body returned as a dict. If
no body was returned the response will be None - location: If a
location header is returned, this value will be the location URL as a
string otherwise it will return None - objectId: If a location
header is returned, the value of objectId will be the last part of the
location url as a string otherwise it will return None - Etag: If a
Etag header is returned, the value of Etag will be the content of the
Etag header returned otherwise it will return None
To output the response in a human readable format when working in an
interactive session use the view_response method:
In [3]: response = client_session.read('vCenterStatus')
In [4]: client_session.view_response(response)
HTTP status code:
200
HTTP Body Content:
{'vcConfigStatus': {'connected': 'true',
'lastInventorySyncTime': '1440445281484'}}If a method needs a URI parameter to work, the NSX RAML Client will compose the URL based on the base URL, parent and child method URL and the supplied URI parameter. To supply a URI parameter, add a URI parameter dict to the call. You can supply multiple URI parameters in the call if needed.
In [5]: response = client_session.read('vdnSegmentPool',
uri_parameters={'segmentPoolId': '2'})
In [6]: client_session.view_response(response)
HTTP status code:
200
HTTP Body Content:
{'segmentRange': {'begin': '5000',
'end': '10000',
'id': '2',
'name': 'legacy'}}If a method supports one or more query parameters, you can supply those optional query parameters in your request, and the NSX RAML Client will add the query parameter for you. To use this pass a query parameter dict to the call:
In [7]: response = client_session.read('nwfabricStatus',
query_parameters_dict={'resource':
'domain-c1632'})
In [8]: client_session.view_response(response)
HTTP status code:
200
.... truncated for brevity ....It is possible to use URI and query parameters concurrently in any call and add as many as the resource specifies.
If a resource requires a body to be supplied with data the body can be composed in the following way:
Check what the body of a call needs to look like by retrieving it out of
the RAML file, and displaying it to stdout using
view_resource_body_example:
In [9]: client_session.view_resource_body_example('logicalSwitches', 'create')
<virtualWireCreateSpec>
<name>mandatory</name>
<description/>
<tenantId>mandatory</tenantId>
<controlPlaneMode>mandatory</controlPlaneMode>
</virtualWireCreateSpec>It is possible to create a template python dictionary using
extract_resource_body_example and display the output structure in a
human readable format to stdout:
In [10]: new_ls = client_session.extract_resource_body_example('logicalSwitches',
'create')
In [11]: client_session.view_body_dict(new_ls)
{'virtualWireCreateSpec': {'controlPlaneMode': 'mandatory',
'description': None,
'name': 'mandatory',
'tenantId': 'mandatory'}}It is possible to change any of the values in the dictionary with the data to be sent to the API:
In [12]: new_ls['virtualWireCreateSpec']['controlPlaneMode'] = 'UNICAST_MODE'
In [13]: new_ls['virtualWireCreateSpec']['name'] = 'TestLogicalSwitch1'
In [14]: new_ls['virtualWireCreateSpec']['tenantId'] = 'Tenant1'
In [15]: client_session.view_body_dict(new_ls)
{'virtualWireCreateSpec': {'controlPlaneMode': 'UNICAST_MODE',
'description': None,
'name': 'TestLogicalSwitch1',
'tenantId': 'Tenant1'}}This example shows how to send the call to the NSX Manager API by supplying the body dictionary in the call:
In [16]: new_ls_response = client_session.create('logicalSwitches',
uri_parameters={'scopeId':
'vdnscope-1'},
request_body_dict=new_ls)
In [17]: client_session.view_response(new_ls_response)
HTTP status code:
201
HTTP location header:
/api/2.0/vdn/virtualwires/virtualwire-1305
NSX Object Id:
virtualwire-1305
HTTP Body Content:
'virtualwire-1305'Some resources in NSX Manager will additionally need the If-match
header. To compose the If-match header, retrieve the content of the
Etag and return it in the If-match header. For example, this is used
in the distributed firewall configuration to deal with conflicts when
multiple users try to concurrently edit rule sets.
This example shows how to retrieve a dfw rule, edit it, and update it via the NSX API:
rule_read_response = client_session.read('dfwL3Rule',
uri_parameters={'sectionId': section_id,
'ruleId': new_rule_id})
updated_rule = l3_dfw_rule_read_response['body']
etag_value = l3_dfw_rule_read_response['Etag']
updated_rule['rule']['name'] = 'UpdatedByRAMLClient'
update_response = client_session.update('dfwL3Rule',
uri_parameters={'sectionId': section_id,
'ruleId': rule_id},
additional_headers={'If-match': etag_value},
request_body_dict=updated_rule)Note that the If-match header is supplied by the
additional_headers dictionary.
Some resources in NSX expect values to be set in XML Tags. This example shows a dfw resource:
In [18]: client_session.view_resource_body_example('dfwL3Rules', 'create')
<rule disabled="false" logged="false">
<name>AddRuleTest</name>
<action>allow</action>
<notes/>
.... truncated for brevity ....The rulehas the Tags disabled and logged. When this type
of Tag is found, it is converted to a key prefixed by @ in the
resulting dictionary:
In [19]: l3rule = client_session.extract_resource_body_example('dfwL3Rules',
'create')
In [20]: client_session.view_body_dict(l3rule)
{'rule': {'@disabled': 'false',
'@logged': 'false',
'action': 'allow',
.... truncated for brevity ....It is possible to set values using the @ prefix, and they will be
converted to a XML Tag of the top level object.
l3section_bdict['section']['rule'][0]['@logged'] = 'true'In some cases NSX uses lists of parameters with repeating keys. For example:
In [21]: client_session.view_resource_body_example('dfwL3Section', 'create')
<section name="Test">
<rule disabled="false" logged="true">
<name/>
<action>ALLOW</action>
<appliedToList>
<appliedTo>
<name/>
<value/>
<type/>
<isValid/>
</appliedTo>
</appliedToList>
<sources excluded="false">
<source>
<name/>
<value/>
<type/>
<isValid/>
</source>
<source>
<name/>
<value/>
<type/>
<isValid/>
</source>
</sources>
<destinations excluded="false">
<destination>
<name/>
<value/>
<type/>
<isValid/>
</destination>
<destination>
<name/>
<value/>
<type/>
<isValid/>
</destination>
</destinations>
<services>
<service>
<destinationPort/>
<protocol/>
<subProtocol/>
</service>
</services>
</rule>
<rule disabled="false" logged="true">
<name/>
<action>DENY</action>
</rule>
</section>There are multiple destination keys under destinations. To be
able to work with python dictionaries, nsxramlclient will convert those
list of equally named parameter 'groups' to a Python list containing
dictionaries. This example shows the resulting Python dictionary for
this type of resource body schema:
In [22]: dfw_l3_sec = client_session.extract_resource_body_example('dfwL3Section',
'create')
In [31]: client_session.view_body_dict(dfw_l3_sec)
{'section': {'@name': 'Test',
'rule': [{'@disabled': 'false',
'@logged': 'true',
'action': 'ALLOW',
'appliedToList': {'appliedTo': {'isValid': None,
'name': None,
'type': None,
'value': None}},
'destinations': {'@excluded': 'false',
'destination': [{'isValid': None,
'name': None,
'type': None,
'value': None},
{'isValid': None,
'name': None,
'type': None,
'value': None}]},
'name': None,
'services': {'service': {'destinationPort': None,
'protocol': None,
'subProtocol': None}},
'sources': {'@excluded': 'false',
'source': [{'isValid': None,
'name': None,
'type': None,
'value': None},
{'isValid': None,
'name': None,
'type': None,
'value': None}]}},
{'@disabled': 'false',
'@logged': 'true',
'action': 'DENY',
'name': None}]}}Note the rule key, its value is a python List containing multiple
rule objects that themselves are python dictionaries. The same holds
true for the destinationsand sources keys.
Copyright © 2015 VMware, Inc. All Rights Reserved.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Any contributions are welcome, bug reports, additional tests, enhancements, etc. Also we welcome your feedback if you find that anything is missing that would make nsxramlclient better