Scripting today is an important skill if you’re a part of IT operations team. It is common to use PowerShell or any other scripting language of your choice to automate repetitive tasks and be efficient in what you do. Another use case for scripting and automation, which is often missed, is the fact that they let you do more. Public APIs offered by many software and hardware solutions let you manipulate their data and call functions in the way you need, without being bound by the workflows provided in GUI.
Recently I was asked to extract data from vRealize Operations Manager that was not available in GUI or a report in the format I needed. At first it looked like a non-trivial task as it required scripting and using REST APIs to pull the data. But after some research it turned out to be much easier than I thought.
Using Python this can be done in a few lines of code using existing Python libraries that do most of the work for you. The goal of this blog post is to show that scripting does not have to be hard and using the right tools for the right job you can get things done in a matter of minutes, not hours or days.
Scenario
To demonstrate an example of using vRealize Operations Manager REST APIs we will retrieve the list of vROps adapters, which vROps uses to pull information from many hardware and software solutions it supports, such as Nimble Storage or Microsoft SQL Server.
vROps APIs are obviously much more powerful than that and you can use the same approach to pull other information such as: active and inactive alerts, performance statistics, recommendations. Full vROps API documentation can be found at https://your-vrops-hostname/suite-api/.
Install Python and Libraries
We will be using two Python libraries: “Requests” to make REST calls and “ElementTree” for XML parsing. ElementTree comes with Python, so we will need to install the Requests package only.
I already made a post here on how to install Python interpreter and Python libraries, so we will dive right into vROps APIs.
Retrieve the List of vROps Adapters
To get the list of all installed vROps adapters we need to make a GET REST call using the “get” method from Requests library:
import requests
from requests.auth import HTTPBasicAuth
akUrl = 'https://vrops/suite-api/api/adapterkinds'
ak = requests.get(akUrl, auth=HTTPBasicAuth('user', 'pass'))
In this code snippet using the “import” command we specify that we are using Requests library, as well as its implementation of basic HTTP authentication. Then we request the list of vROps adapters using the “get” method from Request library, and save the XML response into the “ak” variable. Add “verify=False” to the list of the get call parameters if you struggle with SSL certificate issues.
As a result you will get the full list of vROps adapters in the format similar to the following. So how do we navigate that? Using ElementTree XML library.
Parsing XML Response Sequentially
vRealize Operations Manager returns REST API responses in XML format. ElementTree lets you parse these XML responses to find the data you need, which you can output in a human-readable format, such as CSV and then import into an Excel spreadsheet.
Parsing XML tree requires traversing from top to bottom. You start from the root element:
import xml.etree.ElementTree as ET akRoot = ET.fromstring(ak.content)
Then you can continue by iterating through child elements using nested loops:
for adapter in akRoot:
print adapter.tag, adapter.attrib['key']
for adapterProperty in adapter:
print adapterProperty.name, adapterProperty.text
Childs of <ops:adapter-kinds> are <ops:adapter-kind> elements. Childs of <ops:adapter-kind> elements are <ops:name>, <ops:adapterKindType>, <ops:describeVersion> and <ops:resourceKinds>. So the output of the above code will be:
adapter-kind CITRIXNETSCALER_ADAPTER name Citrix NetScaler Adapter adapterKindType GENERAL describeVersion 1 resourceKinds citrix_netscaler_adapter_instance resourceKinds appliance …
As you could’ve already noticed, all XML elements have tags and can additionally have attributes and associated text. From above example:
- Tags: adapter-kind, name, adapterKindType
- Attribute: key
- Text: Citrix NetScaler Adapter, GENERAL, 1
Finding Interesting Elements
Typically you are looking for specific information and don’t need to traverse the whole XML tree. So instead of walking through the tree sequentially, you can iterate trough interesting elements using the “iterfind” method. For instance if we are looking only for adapter names, the code would look as the following:
ns = {'vrops': 'http://webservice.vmware.com/vRealizeOpsMgr/1.0/'}
for akItem in akRoot.iterfind('vrops:adapter-kind', ns):
akNameItem = akItem.find('vrops:name', ns)
print akNameItem.text
All elements in REST API responses are usually prefixed with a namespace. To avoid using the long XML element names, such as http://webservice.vmware.com/vRealizeOpsMgr/1.0/adapter-kind, ElementTree methods support using namespaces, that can be then passed as a variable, as the “ns” variable in this code snippet.
Resulting output will be similar to:
Citrix NetScaler Adapter Container Dell EMC PowerEdge Dell Storage Adapter EP Ops Adapter F5 BIG-IP Adapter HP Servers Adapter
Additional Information
I intentionally tried to keep this post short to give you all information required to start using Python to parse REST API responses in XML format.
I have written two scripts that are more practical and shared them on my GitHub page here:
- vrops_object_types_1.0.py – extracts adapters, object types and number of objects. Script gives you an idea of what is actually being monitored in vROps, by providing the number of objects you have in your vROps instance for each adapter and object type.
- vrops_alert_definitions_1.0.py – extracts adapters, object types, alert names, criticality and impact. As opposed to the first script, this script provides the list of alerts for each adapter and object type, which is helpful to identify potential alerts that can be triggered in vROps.
Feel free to download these scripts from GitHub and play with them or adapt them according to your needs.
Helpful Links
The main benefit of using Unisphere Domains is that they give you ability to manage all of your VNXs by connecting to just one array. If you have an old Clariion you’ll have to use a so called Multi-Domain. VNX1 and VNX2 arrays can join a single domain.
Yesterday while working on one of the customers’ NetApp array I hit an issue which looked like an SSL misconfiguration at first.
Niktips's Blog 